AI & Machine Learning Solutions for Business

Machine Learning and your Business

We help clients understand how Machine Learning can improve their business.

Types of Analysis

Business analysis currently falls into one of two categories - Business Analytics and Predictive Analytics. Business Analytics is traditional analysis or "reporting". It's data that shows "how many users visited my site?" or "how many sales did we make in July?".

Predictive Analytics relates to machine learning and generates a percentage value representing certainty rather than concrete numbers. It provides results such as "based on last month's sales, how many products will we sell this month?" or "does this image contain a car?".

Business Analysis

Traditional reports that display existing data such as "Total shoe sales in July".

Predictive Analysis

Machine Learning predictions such as "I'm 90% certain this news feed contains angry content"

What is AI, Machine Learning and Deep Learning?

AI, Machine Learning and Deep Learning are all part of the same branch of technlogy, however they do mean different things.

AI means Artificial Intelligence and is the idea of computers thinking, or more accurately, processing data on their own without the need for explicit programming. There are various classifications of AI's however it's safe to say that they all deal with the concept of self-thinking machines.

Machine Learning involes providing an AI large quantities of data in order to teach a concept. A traditional example is showing an AI thousands of cat and dog images and telling the AI which image is which. After it has learned, the AI will be able to identify, with a certain level of accuracy, whether an image is a cat or a dog. The key here is that human involvement is required during the teaching phase. Each image must be labeled properly and depending on the size of the data set this could involve a great deal of human overhead.

Deep Learning is very similar to Machine Learning except that it excludes the human teaching component. With Deep Learning, the AI would input data and output data to teach itself the difference between a cat and a dog.

Is Machine Learning Right for Your Business?

Identify your Machine Learning Needs

Whether or not your business needs machine learning depends on the type of data that you're looking and whether or not certain knowledge could help improve your opperations. If you're only after historical data then you may not need AI. If your business needs to use existing data to make future predictions or identify patterns in data flow, then machine learning can help you.

We often recommend looking at your most labour intensive processes to see if they can be automated by AI. If for example your users upload images to a website and you would like to automatically verify whether or not those images are appropriate, you're a perfect candidate for using AI algorithms.

Take a look at our machine learning Applications below to get an idea of the types of problems AI solve.

Common Applications of Machine Learning

Examples of Machine Learning

Modern machine learning algorithms have a wide range of applications. Here are a few typical ones that are commonly used in business appliations and analytics.

Image Recognition	Identify objects in an image. For example, find images that contain a car, a celebrity or other object.
Suggestive Content	Identify whether an image contains suggestive content. This is execially useful in applications where images are publically uploaded to a system and can be identified quickly by an AI.
Predictions	Using historical data, make predictions about future outcomes. For example, by looking at past sales, predict the total number of sales that will be made this month or get inventory estimates based on trends.
Chatboxes	Conversational interfaces that let users issue commands by speaking at an application or device. Examples of these are Amazon's Alexa, Apple's Siri Microsoft's Cortana and Google's Home.
Entity Analysis	Look through text and identify entities relating to it's content. Entities can be anything such as language used, locations, dates, names, numerical values or names or organizations.
Sentiment Identification	Have your AI parse through data to identify the general sentiment (sad, happy, angry, etc) of the content. This can be useful for quickly identifying and addressing negative content such as angry social posts or comments.
Text to Speech	Text to speech algorithms let text be read out by a computer generated voice. These typically support many languages and voice types.
Translation	Translation services that allow for content to be accurately translated from one language to another. Google's translation AI is famous for inventing it's own intermadiate language in order to make accurate tranlations.

AI & Machine Learning