Who knew that the next time I would be attending a tech conference, I’d be attending as a speaker? Well, neither did I.
GDG DevFests are large, community-run developer events happening around the globe focused on community building and learning about Google’s technologies.
GDG DevFest Nairobi 2017 was an all-day developer conference that offered speaker sessions across multiple product areas, code-labs, hackathon and much more! This years’ DevFest in Nairobi, Kenya was held at Strathmore University on the 11th of November 2017.
One of the co-organizers of WIMLDS (Women in Machine Learning and Data Science), was seeking interested folks to do a beginner Machine Learning code-lab at DevFest Nairobi 2017 and I was happy to take up the challenge. I knew that this would be a great opportunity for me to learn, improve on my communication skills and a step towards becoming a visible expert in my field.
So, my awesome co-speaker Hazel Apondi and I decided to do an introductory presentation on Machine Learning together with a code-lab on sentiment analysis on movie reviews, which we thought would be an interesting topic to discuss and learn in more detail.
While preparing for the presentation I got to learn 2 main things:
1. Why Naive though?
Do you know why the Naive Bayes algorithm is known as Naive? If not, don’t worry, I got you!
In Machine Learning, naive Bayes classifiers are a family of simple probabilistic classifiers based on applying Bayes’ theorem with strong(naive) independence assumptions between the features.
In other words, it assumes that the existence of a certain feature in a class does not depend on other features in the same class. A popular example: A fruit is said to be an apple if it’s let’s say round, red/green and 3 inches in diameter. Even if these features depend on each other or in the existence of other features, it is assumed that all these features independently contribute to the probability of the fruit being an apple and that is why it is known as ‘Naive’. This makes the classifier to perform better compared to other models and you need less training data. You can get more information on Naive Bayes here
2. Jupyter Notebook Slides!
While getting ready for the code-lab and trying to find out the best way to present the code, I came across Jupyter notebook slides! This great program is useful especially for anyone interested in presenting code. I’ve been using Jupyter notebook for a while now but I was unaware that it could do so much more than just writing code. You can check out a post on medium on how to Present code using Jupyter notebook here.
Apart from those two, I also learnt how to do sentiment analysis using the Naive Bayes Classifier, discovered how cool NLTK toolkit is, among many other things. Hoping to be a part of more conferences like this.
You can find our presentations on the links below:
- Power point presentation by Hazel Apondi: Machine Learning Demystified
- Code-lab by Catherine Gitau: Sentiment Anlaysis Code-lab
To more Learning!