Milestone 6.2 - Data science training
Project: AgeingTimeUse, Horizon 2020 ERC
Although I initially thought that I would do multi-level model training, in the course of my project, I realized that I would benefit more from rigorous data science training since I had already mastered MLM when I started the project (see my paper using MLM here).
There are only a few programs that can provide PhD-level data science training. One of them is The Data Incubator. I enrolled there and found out that this was one of the most eye-opening experiences I had had in my academic career. We worked closely with machine-learning and deep-learning models during the intensive course spanning eight weeks. Here is my affiliate link to the program.
The skills I earned as a TDI alumna:
- data wrangling in Python + web-scraping
- Machine Learning (Regressions, Overfitting, Classification, Clustering, Dimensionality Reduction, Scikit-Learn Through and Through, Grid Search and Pipelines)
- Advanced Machine Learning (Naive Bayes, Decision Trees and Random Forests, Time Series, Unbalanced Classes, Sentiment Analysis, Outlier Detection, Recommendation Engines...)
- SQL (believe me, not as easy as it sounds)
- Spark (totally complex, but also awesome)
- Basics of web-development and visualization in Python (Altair)
- Tensorflow and keras (Convolutional NN, RNN, Variational Auto Encoders, Adversarial Noise, Deep NN, DeepDream...)