Share

Artificial intelligence has been around for a while now, from self-driving cars to computers that write movie screenplays, write songs or even voice assistants in our cellphones, AI is starting to rock our world and before we can notice AI will be everywhere.

Every day is easiest for us to use all the potential that can give to our apps and projects. So let’s check how we can use AI and discover that we don’t need to be a genius scientist or a futuristic entrepreneur like Elon Musk to do awesome things with AI.

A really huge tendency now are AI services and bots, with these services and tool you can implement top algorithms and AI features on your apps.

For example, with AI integrated on your app/system you could…

  • Predict a potential paying customer, based on his activities during the first day/week/month.
  • Detect spammers, fake users or bots in your system, based on website activity records.
  • Classify a song genre (rock, blues, metal, etc), based only on signal-level features.
  • Recognize a character from a plain image.
  • Detect, based on accelerometer and gyroscope signals, whether a mobile device is staying still, walking (upstairs or downstairs), lying vertical or horizontal, etc.
  • Extract data from printed documents.

These days every big tech company is developing some kind of product or service related with AI, Google, Facebook, Microsoft, Amazon, Netflix and much more. Some of these companies are making their AI projects and code, open source so everyone could check and reuse the code, but if you’re not really into AI development this is useless.

AWS Machine Learning

“Amazon Machine Learning is a service that makes it easy for developers of all skill levels to use machine learning technology.”

When your developing predictive machine learning software you need to be carefully choosing the model type based on a specific use case. A really cool and powerful feature in Amazon Machine Learning is that you don’t need to care about that, the platform will automatically train and test a variety of complex models, tuned with many different parameters, and will choose the best one for you to make the final evaluation. This cool feature it will save you a lot of time, training ant testing a lot of models and the end you will have the same results in less time.

Another really cool feature it’s the data handling. Amazon ML will make for you the input normalization, data splitting and model evaluations. Again saving you a lot of time.

Obviously Amazon ML will integrate perfectly with AWS cloud platform but not only with AWS Service also will give you a really cool API to work with every single cool feature is explained very detailed in his documentation and the last thing we need to mention about this service is you will pay only for what you use making worth every single penny and giving best suitable scalability models.

More info: http://aws.amazon.com/machine-learning/

Azure Machine Learning:

There are two key features about Azure ML that I need to highlight it’s their really cool user experience and interface and also how easy it’s to integrate with almost every possible type of software solution.

Just like Amazon they have a really good documentation and as soon as you create your first experiment you will realize that working with experiments and linking blocks it’s quite fun and an easy way to work on your models.

Microsoft it’s really pushing this product so they will give many different and interesting pricing options and depending if you’re a visual studio subscriber, BizSpark member or a developer they will give some free features.

More info: https://azure.microsoft.com/en-us/services/machine-learning/

Watson Analytics:

Guided and automated analytics from the cloud. IBMWTSON: http://www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/services-catalog.html The product shined most at data visualization, and while it offered a natural language interface for asking questions about data sets, the only questions reliably answered were of the form "how does X vary with Y", where X and Y are any two attributes in the dataset. the Explore and Refine tools allowed me to dig a little deeper into what I had (by filtering and sampling the data and letting Watson find correlations between columns). This in itself can prove fruitful if we can find a strong, previously unknown correlation that can be used directly. The Predict tool is where Watson starts to get really interesting; allowing the user to use natural language to phase analytical-type questions such as 'what are the top predictors of churn?' or 'does A influence B?' - This is after all what Watson was originally famous for. I really like what IBM is trying to do with Predict, it does take out some of the 'pain' of analytics (like which model types/classes are best for explaining different kinds of target variables and all that tiresome tuning). However, this is the functionality most likely to disappoint those that are hoping Watson is a panacea for self-service, departmental-level analytics;

Google Prediction API:

If Google and especially Amazon have any one guiding tenet to their cloud approaches, it's "less is more." Maybe better to say "just enough is more," which includes the way both companies offer cloud-based machine learning services.

In Google's case, Google Cloud Platform currently offers only two services akin to the others profiled here: Google Translate (an API supporting Google's existing machine translation engine), and Google Prediction API. The former is a proprietary API maintained exclusively by Google. The latter, despite the unassuming name, is a broadly inclusive service that allows users to upload data and train models in the manner of of Microsoft Azure Machine Learning Studio. (Data can be derived from Google services like Google BigQuery.)

Amazon Machine Learning is similar to Google Prediction API in that models can be trained against data and used to make predictions. It's a deliberately simplified service, either for the sake of appealing to developers who only want to solve a specific, narrow problem or because Amazon wanted to test the market waters first.

In both Amazon and Google's cases, their targets are developers both with narrowly defined needs and with data already on those clouds -- the "just enough" model. IBM and Microsoft are aiming for far broader territory, and while IBM strives to have the most to offer, it also has the most to lose.

Other interesting ML services