Watson Studio

This is a “landing page” for Watson Studio. Look here for links to high value technical content specific for this service.

References for Further Reading

• Build a custom model with Watson Visual Recognition Object Detection - nice video (about 9 minutes) showing how you can train a custom model for Watson Visual Recognition in Watson Studio.
• A Visual Intro into NumPy - since we use a lot of Python in Watson Studio, you should be familiar with Python and NumPy is one of the basic Python libraries that EVERYONE uses.
• Watson Machine Learning within the Data Science Experience - a blog post which walks you through setting up a machine learning model for a predictive retail use case. Walks you through the entire process. Some of this is out of date, it mentions the Data Science Experience (DSX), which was the predecessor of Watson Studio.
• Watson Studio Video Library - a HUGE list of videos that will help you with just about anything that you can think of within Watson Studio.
• Data Science Community - a good place to grab tutorials, data sets (yes - real data sets), articles, and notebooks. Nice place to grab things to get you started.
• Python Cheatsheet - this one looks pretty good, if you find others, add the links and let people decide for themselves which one they like.

Python Notebooks

• Get started with Jupyter Notebook and Watson Studio - quick blog post and introduction to Python notebooks in the Watson Studio environment.
• Conversational Assistants and Quality with Watson Assistant — Revisited - nice blog post that highlights some best practices for a deployed chatbot with Watson Assistant. Check out the Python notebook with k-fold testing and intent analysis. Also note the follow up article Conversational Assistants and Quality with Watson Assistant — The Measures, which explains the metrics.
• Watson Assistant Notebooks - some notebooks around testing and Dialog profiling that you can use to evaluate your chatbots.
• Visualizing Geospatial Data in Python - Nice article on using the geopandas and geoplot libraries, to visualize map based data. Really great examples, and a great write up.

Machine Learning

Some of the value that people get from Watson Studio is the ability to create and use machine learning models for a large number of different reasons. Some of this may be for predictive analytics, some for just cognitive analysis, and some for just pure data science.

• An explanation of machine learning models even you could understand - If you are interested in HOW machine learning algorithms work, give this a quick read. You can use AI without knowing some of this “under the covers” stuff - but knowing how the various algorithms work helps with understanding machine learning approaches better.

Data Sets

A lot of machine learning relies on having good data sets. Most applications have their own data, but if you are looking to show a proof of concept, you may not have access to the data that you are looking for. That’s where other data sets come into play.

• UC Irvine Repo - a nice open source collection of data sets from University of California at Irvine.
• Kaggle - Ihave a lot of customers who swear by Kaggle - it’s got some pretty good data sets.