machine learning with Python

I have been learning Data Science with Python since mid 2018, focusing on Machine Learning models. I enjoy the intellectual challenge model building and coding presents and am fascinated to see the vast variety of machine learning use cases. I believe in life-long learning, knowing more now than a year ago and knowing more a year from now than today. On this page I am showcasing some of my scripts, summarize experience and completed courses.

Ready for a bit of fun, solving a puzzle? The following Python code will return my IQ. Still very far from Magnus Carlsen.

import math

IQ = lambda x: ''.join(digit for digit in str(round(x, 4))[-4:-1])

print(IQ(math.pi))

I contributed to a machine learning project at NielsenIQ in 2019 Q2. While of course I cannot share any details, I can mention that the task was to find matching entities in two different data sources based on similarity in behavioral patterns. Technically speaking my Python workflow involved connecting to different data sources (SQL server, csv files, Excel files), getting the right data, data cleaning, feature engineering for a machine learning model, creating datasets for training and testing models, hyperparameter tuning, selecting the best model based on accuracy metrics and saving it for predictions in a later stage.

Out of personal interest I am currently working on a supervised learning model to explain and predict market share performance with marketing drivers like price, distribution level, advertising spend, etc. A further build to this idea is the use of a recurrent neural network to factor in the lagged effect of past performance and past marketing efforts. Please watch this space for updates.