AI News, Machine learning for your business

Machine learning for your business

With all the hype around big data, machine learning, and data science, it’s difficult to see how and where these concepts give your business the advantage.

This blog post is to help you discover and imagine ways to take advantage of machine learning, and data mining in general.

Machine learning in the lab: Machine learning in an operational system: Applying machine learning in an operational system takes some development upfront too.

There are many decisions left to be made for this page. Here are three decisions: One last thing, we need to make the assumption that we already have some data to keep this post short.

In this post we haven’t explained what machine learning is, but hopefully you’ve been exposed to how it can fit inside the decision process for your business.

Machine Learning

Supervised learning algorithms are trained using labeled examples, such as an input where the desired output is known.

The learning algorithm receives a set of inputs along with the corresponding correct outputs, and the algorithm learns by comparing its actual output with correct outputs to find errors.

Through methods like classification, regression, prediction and gradient boosting, supervised learning uses patterns to predict the values of the label on additional unlabeled data.

Popular techniques include self-organizing maps, nearest-neighbor mapping, k-means clustering and singular value decomposition.

How to Apply Machine Learning to Business Problems

It’s easy to see the massive rise in popularity for venture investment, conferences, and business-related queries for “machine learning”

In this article, we’ll break down categories of business problems that are commonly handled by ML, and we’ll also provide actionable advice to begin a ML initiative with the right approach and perspective (even it’s the first such project you’ve undertaken at your company).

For illustrative purposes, it will be helpful to list a number of well-established business use-cases for machine learning so that you (the reader) can churn up your own application ideas:

If you have reams of business data from years ago, it may have no relevance today, particularly in fields where the basic business processes change drastically year-over-year, such as mobile eCommerce).

For example, if you run a door delivery service for pet supplies, and your app, prices, product offerings, and service areas have changed significantly over the last six months, you will need much more recent data to learn from than, say, a company selling homeowners’

While unsupervised learning (see glossary below) allows for a wide degree of applications in making sense of data without labels, it’s usually not advised for companies to “jump into”

One letter difference or one number difference could mean overpaying your bill by 10x the original amount (if the decimal was interpreted to be in the wrong place), or sending money to the wrong company (if an invoicing company name isn’t registered exactly).

As an interesting caveat, there is a San Francisco-based startup called which is aiming to use natural language processing and machine vision to real and pay bills, albeit it pulls humans into the loop before sending funds.

In order to gain additional perspective on the issue of “picking a business problem for machine learning”, we decided to reach out to our network of previous AI podcast interview guests for additional guidance for our business readers: Dr. Ben Waber —

CEO, Calculation Consulting: “The best problems are those in which there is a very large, historical data set that includes both rich features and some kind of direct feedback that can be used to build and algorithm that can be implemented and tested easily and will either decrease operational costs and /or increase revenue immediately.“

CEO, AGI Innovations Inc: (To begin, Peter quotes Dr. Robin Hanson, Professor at George Mason University: “Good CS expert says: Most firms that think they want advanced AI/ML really just need linear regression on cleaned-up data.”) “I think that most businesses don’t justify the investment in ML/DL (of course, ML means many things). Cutting edge stuff that everyone is talking about requires a lot of data and expertise, and is static – i.e.

This leads us into the second major section of this guide: In an off-mic conversation with Dr. Charles Martin (AI consultant in the Bay Area), he mentioned that many companies read about ML with enthusiasm and decide to “find some way to use it.”

Pick a business problem that matters immensely, and seems to have a high likelihood of being solved UBER’s Danny Lange has mentioned from stage that there is one thought process that’s highly likely to yield fruitful machine learning use case ideas: “If we only knew ____.”

could get in the way of developing an effective ML solution: Building a ML solution requires careful thinking and testing in selecting algorithms, selecting data, cleaning data, and testing in a live environment.

Data Science and BD&A, Computer Sciences Corporation: “The most common mistake that businesses make when using ML is that they think that an ML solution is a one-shot process: They send data to data scientists, and data scientists send back THE model.

Log everything, build storage and processing systems, ensure they are accessible, conduct deep analysis and as many experiments as you can on your product, build in intelligence into as much as your product as possible.

The consensus (in the limited number of quotes above, and from dozens of other conversations with business-minded data scientists) is that machine learning is not as much of a mere “tool”

Interested readers might benefit from our recent consensus of 26 machine learning / AI researchers where we asked: “Where should machine learning be applied first in business?” The infographic featured drives home many of the same points highlighted in this article.

The ultimate question for executives remains: When can we have (a) the resources required to invest in machine learning seriously, and (b) a legitimate use case that started from trying to find real business value, not from “trying to find a way to kinda use machine learning.”

6 Examples of AI in Business Intelligence Applications

Companies can now use machines algorithms to identify trends and insights in vast reams of data and make faster decisions that potentially position them to be competitive in real-time.

It’s not a simple process for companies to incorporate machine learning into their existing business intelligence systems, though Skymind CEO and past TechEmergence podcast guest Chris Nicholson advises that it doesn’t have to be daunting. “AI is just a box,” he says.

But as AI has gained momentum, prominent application providers have gone beyond creating traditional software to developing more holistic platforms and solutions that better automate business intelligence and analytics processes.

Each example covers the following: Based on our past interviews withs executives and investors in the field, we predict that business intelligence applications will be one of the fastest growing areas for leveraging AI technology over the next five to 10 years.

HANA takes in information gathered from access points across the business—including mobile and desktop computers, financial transactions, sensors, and equipment at production plants.

If your sales staff uses company smartphones or tablets in the field to record purchase orders, data from those transactions can be analyzed and understood by HANA to spot trends and irregularities.

At a conference hosted by SAP in 2015, then-CIO Karenann Terrell described why Walmart chose to use HANA in order to operate faster and control back office costs by consolidating the processes and resources needed to handle the work.

For example, if a factory manager has an application installed on their computer to monitor the equipment on an assembly line, data from a slowdown in production could be collected and processed through HANA.

It is possible for small and midsize companies, not just enterprises, to explore using this kind of technology in different segments of their business—if the solution can fit within their budget.

The anticipated benefits of using machine learning platforms for business intelligence include infrastructure cost reductions and operational efficiency.

They also projected an average annual benefit of $19.27 million per organization by using HANA, compared with average annual investment of $2.41 million over five years.

Domo, a fast-growing business management software company that’s raised over $500 million in funding, has created a dashboard that gathers information to help companies make decisions.

There are more than 400 native software connectors that let Domo collect data from third-party apps, which can be used to offer insights and give context to business intelligence.

Once these features are rolled out, expected in late spring 2017, the platform is supposed to issue new alerts and notifications for significant changes, such as the detection of anomalies or new patterns in data (similar to approaches used in cyber security already).

Detecting these changes and patterns is expected to fuel the predictive analytics side of Mr. Roboto and help companies predict the return on investment for marketing in real-time, customer churn, and sales forecasts.

“By launching Domo, we were able to quickly optimize and achieve an 80 percent growth in yield during our first quarter,” said David Katz, Univision’s VP general manager for programmatic revenue and operations.

There are numerous ways for machine learning to enhance applications, including those from Apptus, which offer recommendations on actions that companies can take to boost their sales channels.

For example, when a customer visits an online store that uses Apptus eSales and starts to type in search terms to look up products, the machine learning solution can predict and automatically display related search phrases.

Though the technology is still in its early days, Amr Awadallah, founder and CTO at machine learning and software company Cloudera, says deep learning is already adept at prediction and anomaly detection.

According to a study commissioned by Avanade, a survey of 500 business and IT leaders from around the world revealed that they expect to see 33% increases in revenue as a result of smart technologies.

So far, these use cases point to machine learning largely used in service sectors, such as insurance and retail, to address tasks related to customers, sales, and operations;

The increasing prevalence of sensors in machinery, vehicles, production plants, and other hard equipment spaces means physical equipment can be digitized and be monitored by artificial intelligence, a topic we’ve covered before in machine learning applications in industry.

Oil and gas, aviation, and other industries, for example, have been using General Electric’s Predix operating system, which powers industrial apps to process the historic performance data of equipment.

This is a potential threshold moment for business and industry, where machine learning might weave its way further into how operations are handled, the way decisions are made, and resources get managed.

How Effective Managers Use Information Systems

What can managers realistically expect from computers other than a pile of reports a foot deep dumped on their desks every other week?

While there have been advances in basic information retrieval, processing, and display technologies, my recent study of 56 computerized decision support systems confirms the common wisdom that very few management functions have actually been automated to date and all indications are that most cannot be.

In other words, people in a growing number of organizations are using what are often called decision support systems to improve their managerial effectiveness.1 Unfortunately my research also bore out the fact that while more and more practical applications are being developed for the use of decision makers, three sizable stumbling blocks still stand in the way of others who might benefit from them.

Finally, highly innovative systems—the very ones management should find most useful—run a high risk of never being implemented, especially when the impetus for change comes from a source other than the potential user.

Quite simply, my purpose in this article is to discuss, without getting into the technology involved, the high potential of a variety of decision support systems, the challenges and risks they pose to managers and implementers, and a wide range of strategies to meet these challenges and risks.

While there are many ways to categorize computer systems, a practical one is to compare them in terms of what the user does with them: As Exhibit I indicates, EDP reporting systems usually perform only the third function in this list of operations, which I have organized along a dimension from “data-orientation” to “model-orientation.” Hence, unlike the EDP user who receives standard reports on a periodic basis, the decision support system user typically initiates each instance of system usage, either directly or through a staff intermediary.

Incidently, it is interesting to note that external consultants developed the systems cited in my second, fifth, and seventh examples, while those of the first, third, and sixth were the creations of people acting as internal entrepreneurs through staff roles;

In order to help production foremen improve the percentage yield on a newly developed 50—stage process for manufacturing micro-circuits, the management of one company has installed an on-line, shop floor information system.

With this kind of flexibility, the bank’s portfolio managers make more effective use of a vast amount of information, most of which had existed prior to the system, but had been accessible only through tedious manual analysis.

In each case, information extracted from the EDP systems is now maintained separately in order to have it handy and, in two instances, to be able to analyze it in conjunction with externally purchased proprietary data bases and models.

Inputs are projections of future business levels in various lines of insurance and investment areas, plus assumptions concerning important numbers such as future money-market rates.

In order to provide a more rational basis for repetitive marketing decisions, a consumer products company uses a model that relates levels of advertising, promotions, and pricing to levels of sales for a particular brand.

The model was developed in a team setting by reconciling an analysis of historical brand information with an individual’s subjective feelings concerning the effects on sales of various levels and types of advertising and other marketing actions.

Another consumer products company, faced with short-run supply problems for many of its raw materials, has developed an optimization model to solve the mathematical puzzle of choosing and balancing among various product recipes.

As an outgrowth of an overhaul of its group insurance information system, an insurance company has developed a system to eliminate part of the clerical burden associated with renewal underwriting and to help assure that rate calculations are consistent and accurate.

Instead of calculating renewal rates by hand, underwriters fill out coded input sheets for the system, which calculates a renewal rate based on a series of standard statistical and actuarial assumptions.

Although managers in most large companies have used budgeting or planning systems similar to the source-and-application-of-funds model I mentioned, the spectrum of possibilities for other kinds of decision support systems is surprisingly wide.

While many decision support systems share the goals of standard EDP systems, they go further and address other managerial concerns such as improving interpersonal communication, facilitating problem solving, fostering individual learning, and increasing organizational control.

Such systems can affect interpersonal communication in two ways: by providing individuals with tools for persuasion and by providing organizations with a vocabulary and a discipline which facilitates negotiations across subunit boundaries.

At one point, it occurred to the plant manager that he could use this model to investigate whether marketing was setting goals that resulted in poor plant utilization and made him appear inefficient.

As he ran the model under a series of different production mix goals, it became clear that this was the case, and he used the results to persuade marketing to change his plant’s production mix.

With a model that generated optimal training schedules, the scheduler could protect himself very easily by saying: “Using these assumptions concerning attrition, acceptable peak-time shortfalls, and other considerations, this is the best budget.

Decision support systems also help managers negotiate across organizational units by standardizing the mechanics of the process and by providing a common conceptual basis for decision making.

In a number of instances, the development of these definitions and formats was a lengthy and sometimes arduous task that was accomplished gradually over the course of several years, but which was also considered one of the main contributions of the systems.

For example, one of the purposes of some of the model-oriented systems in my sample was to estimate beforehand the overall result of decisions various people were considering separately, by filtering these decisions through a single model.

Although the implementers of a number of the successful systems I studied found it necessary to go through the motions of presenting a cost/benefit rationale which attributed a dollar value to personal effectiveness, they didn’t believe these numbers any more than anyone else did.

Monetary savings are obviously a very important and worthwhile rationale for developing computer systems, but it should be clear at this point that the EDP-style assumption that systems should always be justified in these terms does not suffice in the area of decision support systems.

Again, the general problem here is a common tendency for technical people to concentrate on the “technical beauty” of a system or idea and to assume that nontechnical people will somehow see the light and will be able to figure out how to use the system in solving business problems.

Despite the common wisdom that the needs of users must be considered in developing systems and that users should participate actively in implementing them, the users did not initiate 31 of the 56 systems I studied and did not participate actively in the development of 38 of the 56.

Impose gracefully: Marketing and production managers in a decentralized company did not relish the extra work (format changes and data submission requirements) needed for a yearly budgeting system, which top management was installing.

In another company, management had a real-time system installed for monitoring the largely automatic production of an inexpensive consumer item in order to minimize material loss due to creeping maladjustments in machine settings.

Instead of starting as extensions of existing data processing systems, many decision support systems are built from scratch for the sole purpose of improving or expediting a decision making process.

From Data to Decisions - Machine Learning in Finance

Machine learning is going to change radically the decision making processes in financial institutions. Find out more about Imperial's new Executive Education ...

Machine Learning & Artificial Intelligence: Crash Course Computer Science #34

So we've talked a lot in this series about how computers fetch and display data, but how do they make decisions on this data? From spam filters and self-driving ...

Decision Tree Tutorial in 7 minutes with Decision Tree Analysis & Decision Tree Example (Basic)

Clicked here and OMG wow! I'm SHOCKED how easy.. No wonder others goin crazy sharing this??? Share it with your other friends ..

Decision Making and Reinforcement Learning - Georgia Tech - Machine Learning

Watch on Udacity: Check out the full Advanced Operating Systems course for free ..

Decision Tree Analysis IN HINDI With Solved Practical by JOLLY Coaching

This video is about DECISION TREE ANALYSIS which will help you to understand the basic concept of decision tree analysis. In this video i have solved one ...

The Future of Corporate Finance: Automation powered by SAP Leonardo Machine Learning

You want to leverage next generation automation technologies to significantly increase the level of automation in your Shared Services Organization and ...

Machine learning + neuroscience = biologically feasible computing | Benjamin Migliori | TEDxSanDiego

Whether you're a human, an animal, or a machine, decisions can't be made without perception, which is how we come to understand the world around us.

Machine Learning Techniques and Applications in Finance, Healthcare and Recommendation Systems

David Vogel, Trustee (Voloridge Investment Management, LLC) Abstract: The introductory portion of this talk will review some state-of-the-art machine learning ...

Driving Business Analytics with Data Mining and Machine Learning

Big Data is disrupting enterprises, and Hadoop is helping companies turn this challenge into opportunity. Explore different techniques that allow you to gain ...

Predicting judicial decision-making with machine learning: Carrying on Business

In this webinar, Benjamin Alarie will give a walk-through of Tax Foresight's carrying on business classifier. He will use the classifier on a recently decided case ...