Today’s work brought some success regarding both the improvement of programming skills and methods in data engineering. In the Python classes, I managed to complete two modules – “Booleans and Conditionals” and “Functions and Getting Help”. That increases the total number of completed courses to three out of seven, which greatly enhances my knowledge regarding conditional statements, loops, and modular design. As for machine learning, the team faced a serious problem related to the algorithm, as the fine-tuning of the model gave us poor results – the prediction accuracy was only 34%. This experience taught me a great lesson about data-driven artificial intelligence in that way that the optimization of models is not only based on the architecture but, first of all, on the data itself. Thus, we have understood that the solution of the problem lies in returning to the very basics of the pipeline, namely, to check again the splits between training, testing, and validation data sets. Also, performing three iterations of rehearsals for the weekly presentation made me understand the time allocation correctly.
