What’s a Convolutional Neural Network? A Look on AI’s Promising Tool


The world creates 16.3 zettabytes of data every year. Considering that a single zettabyte is one trillion gigabytes, that’s more than a pool of data. It’s an ocean, and it’s about to get deeper. Research group IDC predicts that the world will be creating 163 zettabytes by 2025. Some of the rise will come from the Internet of Things (IoT), but a large portion will be photos, comments, blog posts, videos, and other forms of unstructured data.

Data scientists are continually developing Artificial Intelligence techniques to analyze this flood of data while it’s still fresh enough to be useful.

One tool showing promise right now is the convolutional neural network (CNN).

Understanding convolutional neural networks requires a little familiarity with the mathematical definitions of the base terms “neural network” and “convolution”.

What’s an Artificial Neural Network?

An artificial neural network (ANN) is a construct used in machine learning which is modeled after the human neural network. It’s built from layers of nodes (called “neurons”) which can perform operations on data received and transmit that data to the next neuron.

Data passes through the layers of the network to produce a final result. The network “learns” to produce increasingly accurate results by passing adjustments back in a process called back-propagation.

What’s a Convolution?

A convolution measures how much one function overlaps another as the first passes over the second. It’s a mathematical way of “blending” the functions and studying the result.

What’s a Convolutional Neural Network?

A convolutional neural network is a specific type of ANN that applies convolutions to the data passing through. CNNs are composed of a convolutional layer (more commonly multiple layers) followed by fully-connected layers, often with intermediate subsampling layers as well.

There are four main layers of a convolutional neural network between the input and output layers.

  • Convolutional Layer (CONV): The convolution is applied to data from the input layer. This layer’s main purpose is to extract features from the input.
  • Activation Layer (RELU usually, but could also be tanh): This layer determines the final value of a neuron by applying a nonlinear activation function to the results of the convolutional layer. All negative values in the matrix are set to 0 while all other values remain constant. Using an activation layer speeds up the training of neural networks.
  • Pooling Layer (POOL): This is the subsampling layer which looks at the max of all previous values. It indicates if a feature was present in the previous layer, but not where, and allows later CONV layers to work on a larger section of the image or data.
  • Fully-Connected Layer (FC; also called the affine layer): Fully-connected layers are connected to all neurons in the previous network, as in a regular neural network. Using a FC layer isn’t mandatory, but it is an easy way to learn a linear function out of the feature space created by the previous layers.

There are very often multiple convolution and subsampling layers, each set looking for a different thing such as:

  • Edges (computer vision)
  • Sound ranges (audio classification)
  • Word grouping (text classification)

Convolutional neural networks are efficient to run and very fast to train. They’re also simpler to set up as they don’t require a large number of defined weights.

Evolving Applications

Computer Vision

Two factors make CNNs especially very well-suited for image recognition: location invariance and local compositionality.

Location invariance means that it doesn’t matter as much where a thing is as that it’s present and recognized.

For example, when sorting pictures of dogs from a group of Facebook images it’s not necessary to know where the dog is within the picture, just that it’s there.

Local compositionality suggests that things which are near each other are often related.

Humans have a very easy time with this, but machines have not so much. By using convolutions to blend features, a CNN can associate features with nearby features for more accurate identification.

Natural Language Processing

It may seem less intuitive to apply convolutional neural networks to natural language processing (NLP), but data scientists are seeing good results using the “Bag of Words” model to represent the text.

In a Bag of Words, individual words are separated and created as JSON objects along with a JavaScript variable representing frequency. The key is the word, the value is the frequency.

Using this model won’t help the convolutional neural network translate a document (other tools such as Recurrent Neural Networks are more useful there).

What they can do is provide a fast solution to document classification and language modeling problems.

CNNs are incredibly fast; they can quickly sort documents into types which can then be processed by slower, more accurate networks.

The Future of Convolutional Neural Networks

CNNs do have some flaws. They need a huge amount of data to process well and tend to learn more slowly than other types of neural networks.

There’s also an issue with translation invariance; convolutional neural networks don’t particularly care if a feature is in the right place, only that it’s present. (There are ways to augment the data to work around this last problem, however.)

Despite the issues, convolutional neural networks are growing in popularity. Their limitations don’t outweigh their value as a fast, reliable way to make sense of unstructured data.

As data scientists develop new ways to draw meaning from spatial relationships, CNNs should become more widely applicable in fields like audio processing and NLP.

Machine learning and neural networks are powering the latest generation of smart, flexible chatbots. To find out how a chatbot can help deliver outstanding customer experiences around the clock, set up a free consultation with one of our experienced developers today!

Request a Consultation

Top 5 CTO Priorities For 2018

Top 5 CTO Priorities For 2018

The way consumers interact with brands has undergone a major change over the past few years.

Consumers now expect personalized service and rigorous protection of their privacy.

Customer experience is becoming the primary differentiator between brands, so it’s more important than ever to provide a smooth experience.

With that in mind, here are the areas where CTOs need to focus in 2018.

1. Secure Transactions

83% of senior executives in a Cisco security survey were more concerned about digital security now than they were 3 years ago, and for good reason.

Increased global connectivity has created more opportunities for digital theft. The average cost of a single data breach is $3.62 million this year.

By 2019 global data breaches will cause $2.1 trillion in losses.

With expanded personal data sharing due to the growing Internet of Things trend, experts have warned that could rise sharply if transaction security isn’t made a priority.

That means focusing on PCI compliance, using AI-powered fraud detection and antivirus software, and using security-conscious features like SSL.

CTOs also need to resist the urge to take shortcuts on development; many security weaknesses can be traced to poorly-done programming shortcuts.

2. Expanded Payment Options

Technology has created a flood of new payment methods designed to make payments convenient and secure.

Shoppers can check out with PayPal both online and at a growing number of brick and mortar locations.

Phone-based mobile wallets like Apple Pay, Android Pay, and Samsung Pay have seen a rise in usage as well.

Consumers are responding favorably: as many as 54% of Americans have used a digital payment method of some kind.

They enjoy the convenience and security offered by using a trusted service instead of carrying around credit cards.

Offering a selection of payment methods helps prevent churn, as well: a fifth of customers fall out of the buying cycle at the checkout because they don’t see a payment option they trust.

3. Multi-channel Presence

Impulse-buying beyond the small-ticket stage is becoming a thing of the past.

Customers generally browse using different devices and platforms before making a purchase.

Because customers trust each other and online influencers more than ads, this could means asking for advice on social media.

Other times it involves checking online reviews, even when shopping in person.

42% of customers read reviews on products using a mobile device while in a physical store.

CTOs can take advantage of this trend by building a strong, consistent multi-channel presence across their website, social media pages, and mobile apps.

Encouraging user-generated content (like reviews and photos) is key. It has two main benefits:

  • Improved customer trust: When a company welcomes and responds to feedback- even criticism- it creates an impression of transparency and reliability.
  • Higher sales: Customers are much more likely to buy after interacting with user-generated content.

4. Enhanced Mobile Experience

Mobile is becoming a huge part of how customers shop. It accounts for a fifth of e-commerce, and the mobile e-sales growth rate is double that of desktop rates (59% vs 17%).

CTOs need to ensure a smooth, reliable shopping and buying experience.

That means skipping complex interfaces for easy-to-navigate menus and optimizing for mobile performance and security.

Another option to consider is a storefront app.

Storefront apps have 40% higher conversion rates than mobile websites.

They offer better performance, too. The improved experience leads to longer shopping sessions and higher customer satisfaction rates.

5. Artificial Intelligence

Customers want customized service, and AI is the tool that will help CTOs provide that service in 2018.

Bring AI into customer-facing technology stack through:

  • Chatbots: 56% of customers would rather resolve an issue through chat than over the phone. Using chatbots, companies can be responsive to customer needs around the clock in the format they prefer.
  • Personalized upsells: AI algorithms review an online shopping cart and suggest related items at the checkout. Intelligent upsells increase a customer’s potential value by 10-30%.
  • Relevant push notifications: Targeted push notifications have four times the response rate of mass notifications.

Are you ready to revamp your company’s digital presence? Concepta can show you how to take advantage of these trends to make 2018 your most profitable year. Contact us for your free consultation!

Request a Consultation

Download FREE AI White Paper

2018 Technology Predictions for Enterprise

2018 enterprise technology predictions

Staying on top in the tech world means constantly looking ahead.

It’s easier, cheaper, and more effective to integrate new technologies while building applications instead of trying to work them in later.

To that end, here’s a look at where enterprise technology is headed in 2018.

Narrow Artificial Intelligence shows ROI

Scientists have been enthusiastic about artificial intelligence for years, but excitement is spreading to the business community as it demonstrates real world value.

Artificial intelligence can generally be defined as either “general” and “narrow.”

General AI describes machines that can reason and make decisions like a human, without a set domain or task to focus on. It can shift from one task to another as easily as humans.

General AI is the “Holy Grail” for computer scientists, but no one has managed to achieve it yet.

It doesn’t help that every time technology reaches what was once considered general AI scientists realize it’s not comprehensive enough.

That task is then relegated to the narrow AI category.

That doesn’t make them useless, however.

Narrow AI applications are very competent within a single domain or when performing a set task.

That may sound limiting, but in reality it’s the only kind of AI showing ROI.

Every AI currently seeing regular enterprise use is narrow AI:

Computer vision and natural language processing (NLP) are also considered narrow AI in their present forms, though they will likely be folded into general AI eventually.

The use of narrow AI will expand dramatically in 2018.

Major companies are investing in AI, like Ford’s $1 billion Argo AI project aimed at developing self-driving car brains.

Cloud computing will push AI into the mobile realm, too.

79% of tech leaders say increasing AI usage is a priority, and 40% of digital transformation initiatives this year will include AI.

Voice and Visual Search become dominant

Technically these are both forms of narrow AI, but they’ve seen such a surge in interest that they deserve special notice.

Visual search uses an image as the basis for a query instead of text. Users can search for other versions of the same picture, similar pictures, or general information about the image.

As computer vision improves visual search is becoming more viable. While it once only let users find the origin of images, it now offers more options:

  • Find photos of friends across the internet
  • Pull up order pages for items worn by celebrities
  • Get recipes for new favorite dishes
  • Research vacation locations

Visual search is already seeing expanded use by major companies.

Google is fine-tuning its new Google Lens, which lets users take a picture with their cell phone and execute a search based on objects within the frame.

Pinterest has a Lens Your Look feature where pinners can take photos of their clothing to get style suggestions. T

hey’re also upgrading their image search to a Responsive Visual Search, where users have the option to specify what interests them within a photo.

Voice Search allows for search based on spoken commands. In the past voice search was so unreliable it inspired jokes on late night TV, but it’s improved a great deal.

In 2012 the average error rate for voice search was 20%. This year that number fell to 8%.

As the error rate has dropped, people are finding it much easier and more reliable to use. 87% of consumers think voice search is accurate enough to use.

21% of mobile users activate their voice search daily, with half using it at least once a week.

The main draw seems to be the ability to safely use smartphones while driving: 53% of those who use voice search regularly do so behind the wheel.

Because of the rising appeal of these features, expect heavy investment in computer vision and natural language processing throughout 2018.

2018 enterprise technology predictions infographic

Blockchain edges into more common enterprise usage

Blockchain is a secure form of distributed digital ledger consisting of a “chain” of individually encrypted “blocks.”

It’s managed via a Peer to Peer (P2P) Network; everyone has a copy of the chain, which makes altering or forging blocks impossible.

Blockchain is famously used for cryptocurrencies like Bitcoin or Ethereum.

There are more uses for blockchain than tracking cryptocurrencies, though.

Storing information in a blockchain is like storing it in a shared database which is being constantly reconciled.

Records are public and easily verifiable.

Once a transaction is recorded it’s incorporated into the chain, not easily erasable by any stretch of the imagination.

Blockchain also exists on a distributed network, so can’t be censored or altered.

The secure nature of blockchain ledgers has many enterprise applications:

  • Smart contracts: Digital agreements can be structured with blockchain to be fulfilled automatically when specific conditions are met. AIG is testing smart contracts for international insurance policies.
  • Preventing voter fraud: Each vote can be accounted for securely and transparently. This hold enormous appeal for civil rights groups looking for ways to prevent rigged elections.
  • Secure resumes: Create an unalterable CV that lets HR managers trust international hires. Secure resumes are especially helpful for jobs with security concerns.
  • International payment systems: Blockchain provides a method for making payments around the world without worrying about embezzlement or funds being intercepted.
  • Transparent supply chains: Secure tracking eliminates confusion and highlights trouble points in the supply process. It could improve global trade by instituting a level of accountability and trust between trade partners.
  • Secure health records: Hospitals can provide better care and cut down on waste and abuse with an unalterable medical record.

Blockchain technologies are growing in number and sophistication.

More than 2500 new blockchain patents have been filed over the last year.

Analysts predict that blockchain will disrupt the insurance and banking industries in particular.

Augmented Reality continues to mature

Augmented reality involves laying computer-generated or animated assets over a camera image. It’s already seen popular use in entertainment:

  • Pokémon Go
  • Snapchat filters
  • Star Wars Find The Force mobile app

This year, technological advances are pushing augmented reality beyond gaming. AR will find a home in other industries.

  • Retail: RayBan and tattoo shop Inkhunter both have “try on” apps that utilize AR.
  • Education: AR maps constellations over a live camera image of the sky or highlight geological features in real time.
  • Travel: AR gives translation apps a boost by overlaying translations on signs or menus.

Investment in augmented reality is rising.

Spending will double in 2018, going from $9.1 billion to $17.8 billion by the end of 2018.

Commercial use will make up 60% of that spending.

Some analysts suggest that 85% of all AR investments will be commercial (as opposed to games) by 2020.

Google, Apple and others are planning AR glasses, but mobile devices are the priority for 2018.

They offer a wider audience with lower investment costs.

Competition intensifies for technical talent

As companies push digital transformation, the talent gap will be more keenly felt.

There are 3 million more STEM positions than available workers right now.

40% of companies looking for tech talent aren’t in the tech industries, just forward-thinking companies looking for help with digital transformation.

The fields in highest demand are data science, business intelligence, digital security, and cloud computing.

There’s also a call for experienced project managers.

65% of tech leaders say that hiring shortages are holding their digital development back. That’s up from 59% last year.

How are they getting around the skills gap? 60% use third parties to support organizational bandwidth. 55% are outsourcing some or all of their analytics needs.

Expect fierce competition for experienced tech talent in 2018.

Nurturing loyalty in employees with tech skills and offering attractive retention packages will become critical to prevent losing staff to headhunters.

Predictive analytics will drive a “quantifiable” model of business processes as Big Data investments rise

Global enterprise investments in data and analytics will surpass $200 billion a year by 2020.

Companies will be looking for ways to quantify business functions where possible in order to get the best possible ROI from these investments

There is a proven advantage to pushing predictive analytics.

48.4% reported measurable results from Big Data investments.

This year, they want to spread that success throughout their organizations by setting quantifiable goals and enacting data-driven operational programs.

Application security focuses on being resilient, becoming stronger from attacks

Security threats have been a major sore point in 2017.

Ransomware attacks rose sharply. Shadow IT and casual “bring your own device” (BYOD) policies exposed companies to hacking and theft.

There was an overall 164% increase in breached records last year. Not only big corporations get hit: 43% of cyber attacks target small to medium businesses.

To meet evolving threats, security software itself needs to evolve.

The focus is shifting from brute strength protection to intelligent detection and response.

Adaptive security software should continuously monitor and scan for threats and address issues on its own.

Once the problem has been addressed, the software should make changes to fix the weak spots exposed by the attack.

That flexibility is where security companies will focus moving forward.

Security needs to be built into software from the beginning, so 2018 will see increasing collaboration between security personnel and programmers and project managers, as well.

Automation will give businesses a competitive edge in the global marketplace

Between increasing amounts of data, a fast-paced global economy, and a tech skills shortage, companies need strategies to do more with less resources.

Automation meets that need by taking tedious or routine tasks off human hands.

Highly automated companies are six times more likely to experience revenue growth of more than 15%.

Companies are responding to such a clear demonstration of potential: 54% of global companies use automation technologies, and more plan to do so in 2018.

The biggest increase will be IT and data management.

Automation will also shorten software development cycles without decreasing quality (through automated testing)

Where will 2018 take your company? If you’re looking to push digital transformation, Concepta can show you the way! Our developers have the right skills to help you visualize your data and unite your digital operations. Schedule your free consultation today!

Request a Consultation

Download FREE AI White Paper

How Will Face ID Be Used for Attention-Tracking in Ads?

face id ads

While fans are excited about the new iPhone X, there are a number of people questioning how Apple plans to incorporate the advanced Face ID technology into their advertising programs.

Read on for an overview of Face ID, its enterprise applications, and the safeguards Apple has in place to protect users.

What is Face ID?

Face ID is a biometric authentication process that uses a TrueDepth camera system to project an invisible net of dots onto the user’s skin.

An advanced algorithm uses the dots to map the geography of their face, creating a 3D model for later comparison.

Unlike previous facial recognition programs, Face ID adapts to changes in a user’s appearance.

It updates its model with every check-in to account for factors like facial hair, weight fluctuations, glasses, hats, and more.

Is it safe?

Apple’s engineers say identity theft is highly unlikely with Face ID.

There are a number of barriers to “facial theft”, the foremost being that a user’s direct biometric data never leaves their personal device.

Even the mathematical facial model used for Face ID is encrypted within the device’s Secure Enclave.

While third parties can offer Face ID as a form of authentication, they won’t have access to the user’s actual facial model.

They will only receive a binary approval or rejection from Face ID.

Moreover, every app must ask for permission to access Face ID.

It remains to be seen how specific that request must be (will Face ID cover attention tracking as well?) but at the very least, users can decide for themselves how much to share.

That puts control over privacy in the hands of consumers.

How does attention tracking work?

Attention tracking has been part of the Apple code since iOS5, but it wasn’t useful until neural networks matured enough to handle matching and detect spoofing.

Now the technology has made a host of new features available for consumers and app developers alike.

Face ID’s attention tracking algorithm first uses the device’s camera to mark the positions of a user’s eyes and mouth, then calculates the angle from each feature to the phone.

That data predicts whether the user is looking at the screen and where on a page their attention is focused.

Users can access some “attention aware” features without even enrolling in Face ID.

The iPhone X automatically dims or disables the display when users look away.

It also lowers the alert volume when users are actively engaged with the phone, signalling the alert with a visual cue instead.

On the creator’s side, the possibilities are even more exciting.

Developers can use Face ID in conjunction with Apple’s suite of Augmented Reality development tools, called “ARKit”, to built AR apps that integrate a user’s facial data into their function.

According to the press release apps will be able to detect the “position, topology, and expression of the user’s face, all with high accuracy and in real time.”

What business advantage does attention tracking offer?

Detecting a user’s expression and being able to relate that to what they’re looking at is a major clue to how they feel about content.

Face ID could provide real-time performance metrics for entertainment, advertising campaigns, and more.

Apple’s developer policy forbids the use of facial model data for marketing purposes.

Attention tracking isn’t specifically included in that restriction, though.

The policy refers to the mathematical models used for identification.

This opens up attention tracking as an innovative method of judging public response to digital marketing campaigns and building accurate customer profiles.

For more info, read our blog post Computer Vision: Facial Recognition for Businesses.

Looking forward

The iPhone X went on sale in early November, so it’s still too soon to measure Face ID’s lasting impact.

It will be interesting to see how users respond to ARKit integrations and attention tracking in the long run.

If this takes off, it could be a major step forward in improving the customer experience.

Are you curious about Face ID? Concepta’s custom programming team can walk you through the basics and decide if Face ID is right for your next app. Reach out now for your free consultation!

Request a Consultation

Download FREE AI White Paper

Machine Learning and Artificial Intelligence: What’s the difference?

difference between machine learning and artificial intelligence

Machine learning has fundamentally changed the way computers work.

Because machine learning is one of the most well-known evolutions of artificial intelligence, it’s natural that the two terms are often used interchangeably by non-engineers.

However, they aren’t quite the same thing.

The distinction is mainly hierarchical, but keeping it in mind helps to better understand the practical applications of both.

Read on for a discussion of the differences between these concepts and a glimpse of what’s coming in the future.

What is Artificial Intelligence?

Artificial intelligence is one of the fastest growing yet least understood tech trends.

Some people fear AI will take jobs from humans or lead to Hollywood-style robot wars.

The reality is that artificial intelligence is both less fanciful and more intriguing than fiction suggests.

Instead of rendering human workers obsolete, it gives them tools to become more effective and frees them to pursue more highly skilled tasks.

The concept is very broad.

Artificial intelligence strives to create machines that can behave in “intelligent” ways.

Given a large subset of data, AI could make its own decisions about relevance and priority rather than relying on subroutines.

AI-driven processes don’t need predetermined guidelines for every possible situation.

They have the ability to judge situations and take the most reasonable action without needing human oversight.

This is a major departure from non-AI programs.

Even the most highly refined logical algorithm can’t account for the millions of tiny elements involved in everyday tasks.

Consider email sorting.

There are very sophisticated algorithms used in evaluating whether a particular email matters to the account holder, yet dozens of mass advertising messages find their way to the inbox.

Too many variables affect the outcome: sender, content, past interactions, and even date. In its ideal state, artificial intelligence could scan a message’s content and combine that with metadata and other factors to create a “living inbox” where the most relevant emails are always listed first.

AI was divided into specific subfields for most of its history.

Each application was treated as a different subject, and there was little interaction between subdisciplines.

These fields covered diverse topics such as:

  • Language processing: Understanding human language
  • Computer vision: Identifying and handling unlabelled images
  • Genetic algorithms: A method for solving optimization problems using the same process evolution
  • Decision theory: The study of how decisions are made and why

Game Changing Shift in Focus

Machine learning is a subdiscipline of artificial intelligence that aims to give machines the ability to learn from previous experiences and use that knowledge in future interactions.

Essentially, a computer is given a pile of data and a machine learning algorithm to process it.

The algorithm sorts the data, adjusting itself after mistakes, until it can achieve the desired results with a high degree of accuracy.

Machine learning does require a data scientist to adjust the model, choose the algorithm, and subset the data.

Still, it represents a huge leap forward in teaching machines to think.

The advent of machine learning was a unifying force on AI research.

It’s wide applicability meant it could be used in many fields, and learning has become a much-desired characteristic of artificial intelligence.

There are two major types of machine learning.

In supervised learning, the algorithm begins with a training dataset of labelled input variables and output variables.

An algorithm is used to map the relationship between variables so that for any new input variable, the correct output can be predicted.

“Supervised” refers to how the training dataset is like a teacher checking the algorithm’s answers against itself.

Unsupervised learning deals with data that doesn’t come with a labelled set of outcomes for reference.

It finds patterns among blocks of data.

Cluster modelling, the most widely used unsupervised learning method, groups data points by their most relevant traits.

It’s often applied to sales data during the customer classification and segmentation process.

Other unsupervised methods include:

  • Pattern mining
  • Data mining
  • Image and object recognition
  • Sequence analysis

Machine learning in action

Machine learning has an astounding variety of end applications.

There are too many to describe them all, but they can be broken down into a few functions.

Distinguishing relevant features (Classification):

Machine learning finds patterns within data as well as areas where there are no consistent similarities.

These patterns informs an assessment of the relative importance of the data.

It used to take years for a human worker to sort and identify the relevant features of a disordered dataset.

Machine learning “shakes out” these features in a fraction of that time.

Recognizing trends:

Machine learning excels at recognizing trends in data based on relevant features.

It predicts the classification of incoming data according to past outcomes.

This method- using a model to predict future events- is called time series forecasting, and it has powerful implications for business.

Companies can use insight gained through machine learning to prepare for future disruptions, adjust their supply chain in response to anticipated increases in demand, and decide where to focus new campaigns.

Model selection/Fine-tuning parameters:

For any given artificial intelligence process there are millions (sometimes billions) of factors that affect the process’ operation.

Small changes in these factors can increase or reduce the accuracy of an algorithm’s outcome.

There are too many for a human to manually adjust.

Trying to choose the perfect setting for each would take years.

Machine learning techniques can be used to find the optimal setting for each involved variable.

The Limitations of Machine Learning

Machine Learning isn’t a perfect solution for every problem, of course.

As game-changing as the technology is, it hasn’t advanced to a fully autonomous level yet.

There are limitations to how it can be used.

  • Machine learning requires a lot of data.Its nature means that machine learning works best on vast amounts of data.The more data is fed through the algorithm, the more refined it will become.That leads to faster processing and higher accuracy.

    Gathering and structuring enough of the right sort of data could present a challenge; at least half of a data scientist’s time is preparing data for machine learning.

    This is more of a statistics problem than a machine learning problem, and there’s a lot of labelled training data available for most purposes.

  • Most algorithms need to be trained for their intended use.With the exception of Neural Networks and similarly versatile examples, machine learning algorithms have to be directed to a specific application.While the core model may be reusable, experience gained in filtering spam isn’t very useful for image clustering.Refining an algorithm takes time, too.

    Machine learning requires lengthy offline training before reaching the point where it adds value.

  • Machine learning systems are hard to test and debug.To describe machine learning as complicated would be a massive understatement.As a consequence machine learning systems hard to assess and maintain.Traditional software can be tested for functionality using Boolean-based logic (“This program works as expected”), but engineers use degrees of success when evaluating machine learning (“This algorithm produced 85% accurate results and has improved from the last test by 10%”).

    As an interesting wrinkle, it isn’t always possible to be absolutely sure whether machine learning has produced the “correct” result.

    Its results are often more indicative of what most people would say rather than what is actually true.

    Google’s Director of Research Peter Norvig explains the dilemma: “For some problems, we just don’t know what the truth is.

    So, how do you train a machine-learning algorithm on data for which there are no set results?”

What else is out there?

It’s hard to draw a line between machine learning and other artificial intelligence fields like computer vision or natural language processing.

Machine learning has become such a useful way of approaching AI that it’s often incorporated into other applications.

Essentially, artificial intelligence systems that can learn from their mistakes and new data involve machine learning.

There are systems that exhibit “intelligence” without learning on their own.

An example would be an expert system – software programmed to function as an expert in a specific domain.

Expert systems use rules, probabilistic reasoning, and logic to reach conclusions rather than relying on past experience.

They’re capable of providing advice, solving problems, demonstrating processes and explaining their logic, and predicting results.

They have trouble working around gaps in its knowledge base, however, and don’t learn or refine themselves.

The Future of Machine Learning

Just as machine learning is the natural successor to artificial intelligence, Deep Learning is on the cutting edge of machine learning.

It’s the next logical step.

Deep Learning deals with neural networks – algorithms designed to mimic the function of the human brain.

These aren’t the primitive neural networks of the 90s, though.

Scale is of paramount importance.

Deep Learning neural networks are huge, fed with as much data and spread across as many machines as possible.

The more layers and data incorporated into the network, the more accurate the results.

There’s a lot of hardware and training time involved in bringing them to a functional maturity.

In essence, Deep Learning is machine learning on an epic scale.

Final notes

Distinguishing artificial intelligence from machine learning is like differentiating between automobiles and electric cars.

It’s a matter of succession and inclusivity.

In other words: all machine learning is artificial intelligence, but not all artificial intelligence is machine learning.

What are you doing with all your data? Talk to Concepta about building AI applications that will give your company a competitive edge.

Request a Consultation

Download FREE AI White Paper

How Digital Transformation Is Impacting Small and Medium Businesses

digital transformation impacting SMBs

The digital revolution is disrupting the traditional business model for small and medium businesses (SMBs).

On one hand it makes it possible for them to compete with much larger companies, but on the other the investment required can be daunting.

Before setting out to create a digital strategy, it helps to work through what digital transformation actually means and how that affects SMBs.

Changing the Pace of Business

Digital transformation is just that- a total restructuring of operations to function more efficiently in the digital era.

Its effects reach into every area of business.

Here are a few of the core components involved in a successful transformation:

Mobile presence: SMBs used to be fine with a regular website, but now consumers expect a fast, fluid mobile experience.

This is doubly urgent for small businesses.

40% of all mobile searches are for local businesses, and 88% of people who search for a local business will visit either it or a competitor within 24 hours.

60% won’t visit or recommend a business after having trouble with a poorly designed mobile site.

Optimizing for mobile is obviously important- yet 47% of SMBs still don’t have a mobile-friendly website or app.

Enterprise apps: Enterprise apps rework everyday functions- ordering, reporting, marketing, planning- into streamlined processes that can be managed via user-friendly apps.

They cut down on internal confusion since everyone has the most up-to-date information in the palms of their hands.

Implementing enterprise apps at the operational level eliminate redundant operations, increase efficiency, and improves employee morale.

A study by Adobe found that investing in enterprise apps netted companies a 35% ROI on average.

Chatbots: Chatbots are the answer to a major customer dilemma: how can a company provide twenty-four hour customer service at scale without the expense of hiring and training humans agents?

Natural language processing technology has advanced enough that chatbots can handle the majority of routine customer queries.

Technology experts predict that customers will conduct 85% of their brand interactions without speaking to a human at all!

Automation: Digital revolution creates a lot of work, but fortunately much of it can be automated.

Automation involves creating a list of processes that trigger other actions or processes without needing to check with a human first.

One example would be generating order confirmation emails after a customer completes a purchase online.

Automation can also walk a customer through basic troubleshooting after an error is reported.

It’s often confused with artificial intelligence, though automation uses set rules to complete its tasks instead of analytical reasoning.

AI-powered customer management: There’s a marketing adage that 80% of business comes from 20% of customers.

With artificial intelligence, that’s changing forever.

Intelligent profiles formed by automatic customer classification and segmentation help companies identify their best customers.

These profiles also help provide the kind of personalized experience that keeps customers happy and loyal.

Data science and analytics: In many ways, data science is the cornerstone of digital revolution.

Data has been called the new oil, and for good reason.

Increasing data utilization is one of the fastest ways to grow a company, resulting in lower operating expenses and higher revenue.

For more about data science and its impact on the business world, read our white paper: “How Businesses Can Use Data Science and AI to Gain a Competitive Edge.”

The Looming Threat to SMBs

When unchallenged, large companies who refine their digital strategy can satisfy needs once available only through a SMB.

The traditional advantages of small businesses over corporations are personalized service and an inventory of niche products tailored to their local market.

Techniques like intelligent customer profiling give companies that would ordinarily be too large to customize their offerings the insight to do so.

If SMBs aren’t pushing digital transformation themselves, these large companies could steal their client base and push them out of a local market.

The problem is the investment in time and resources required for traditional digital solutions.

Getting maximum efficiency from daily operations like reporting, inventory, or accounting is easy to do with modern software.

However, the type of programs used by corporations aren’t practical for SMBs.

They’re complicated to operate, needing trained staff to maintain their databases, and the typical SMB will only use a fraction of their capabilities.

To complicate the issue, large-scale software is expensive enough that recouping an investment would take too long to merit the expense.

SMB managers often make do by stretching the capabilities of programs like Excel, but that can cause more problems than it solves.

Analytics software inspires a similar dilemma since the rewards of data science and analytics for SMBs are hard to see.

Unsure how to translate their data into actionable results, owners hesitate to invest the time and money to join the digital revolution.

They worry that the potential damages from project failure are much higher for smaller businesses that can’t absorb losses like huge companies.

The Path to Digital Adoption

Fortunately, software developers are beginning to cater to the digital transformation needs of SMBs.

Solutions aimed at SMBs are more widely available.

Owners no longer have to buy management software meant for global corporations in order to digitize their operational needs.

Instead they can choose software with only the features they will use.

For example, a landscaping company can have a unified dispatch and reporting app built that lets managers assign jobs and receive completion reports.

The app costs less than a solution meant for larger companies and meets the company’s requirements more closely.

Analytics is easier now, too.

Rather than hiring an in-house data science team, SMBs can take advantage of off the shelf enterprise analytics software.

This is a popular option among SMB owners.

Last year SMBs used an average of 4.8 apps to manage their operations, up from 3.8 in 2015.

46% are tracking their social media metrics through analytics programs, and 47% use some level of business intelligence software.

SMBs do run into trouble with premade software that doesn’t quite meet their needs.

Some solve the problem by stringing together a collection of apps that each solve a different problem.

The resulting technological complexity can give the impression that data science is too complicated for SMBs, but there are good alternatives to the “patchwork app” system.

Custom programming on an SMB level is surprisingly affordable.

Instead of using a handful of apps to manage outcall scheduling and reporting, for example, a customized business app could combine those functions in one easy to navigate place.

Cloud technologies are making many of the same analytics favored by large companies available to SMBs, too.

By tracking and predicting customer needs, SMBs can implement smart inventory systems that make the most of limited shelf space.

Targeted marketing is another convenient tool for reducing operating costs.

It lowers the price of customer acquisition while raising the value of individual clients through repeat business.

Speaking of lowering costs, data science can reduce overhead in general.

Artificial intelligence and automation takes tedious or repetitive tasks out of human hands, leaving employees free for more skilled projects.

The boost in efficiency makes up for SMB’s comparatively smaller staffs.

Leveling the Playing Field

In a very real sense, SMBs are better positioned to benefit from digital transformation than large companies.

41% of SMBs feel their size is an advantage when overcoming institutional resistance to adopting new technology.

They have less bureaucracy surrounding the decision to change, and they have more to gain by going digital.

When they do commit to digitization, their efforts have a high success rate.

Three quarters of SMBs feel that gains from investing in data science technology met or exceeded their original expectations.

Despite the challenges, digital strategy should be a priority for SMBs.

It’s a game-changer.

Half of industry leaders believe that technology levels the playing field between small businesses and large corporations.

Digital transformation is the most reliable path to maximizing an SMB’s resources to gain an edge against their bigger competitors.

How has the digital revolution affected your business? For advice on fine-tuning your digital strategy or to explore how to begin, get your free consultation with a Concepta expert today.

Request a Consultation

Download FREE AI White Paper