Welcome!

My name is Lovkush Agarwal. I recently decided to change careers and become a data scientist. Following David Robinson’s’ advice, I decided to create this blog, to record my progress, learning and projects.

Posts

• EuroPython Conference 2020, Day 2

  Summaries of the talks and my thoughts for Day 2

  Jul 24, 2020

• EuroPython Conference 2020, Day 1

  I am attending my first ever tech-related conference. Here I record my thoughts on Day I.

  Jul 23, 2020

• Santander Dataset, Part III, Learning from others

  I end this series by describing what I learnt by reading other people's kernals on Kaggle.

  Jul 18, 2020

• Neural Networks, Part II, First MNIST model

  I use the vanilla network from the first part on the MNIST data, achieving an accuracy of 94.9%.

  Jul 14, 2020

• Santander Dataset, Part II, Feature Selection

  I carry out several feature selection algorithms, with the hope of removing features that are reducing the performance of the models.

  Jul 13, 2020

• Neural Networks, Part I, Basic network from scratch

  I create my first ever neural network. It is a vanilla network, written from scratch in Python.

  Jul 9, 2020

• Santander Dataset, Part I

  I start a new project modelling another Kaggle dataset. To start things off, I create some default models, to establish a starting point for future models.

  Jul 1, 2020

• Investigating Credit Card Fraud, Part VI, Summary and Lessons from Kaggle

  I end this project by summarising what I did and summarising what I learnt by having a look at other people's examples on Kaggle.

  Jun 25, 2020

• Stop and Search, Part III, Data Analysis

  I finish this project by plotting various charts to summarise the data obtained in the previous two parts.

  Jun 22, 2020

• Stop and Search, Part II, Data Cleaning

  In this post, I describe the cleaning I did on the data.

  Jun 17, 2020

• Do students do their homework last minute?

  During my previous job teaching mathematics at the University of Leicester, I did a project to investigate whether students did their homework last minute.

  Jun 16, 2020

• Stop and Search, Part I, Data Collection

  In light of the current prominence of BlackLivesMatter, I decided to investigate crime in relation to race. Here I describe how I collected the data I will be analysing.

  Jun 15, 2020

• AIs for Games, Part III, Pruning Min-Max for Pentago

  I describe the various ways I made the algorithm from Part II more efficient. These resulted in big improvements in the efficiency.

  Jun 9, 2020

• AIs for Games, Part II, Min-max for Pentago

  I created an algorithm to search through the game-tree of pentago to a given maximum depth. The codes works, but it is highly inefficient.

  Jun 4, 2020

• Investigating Credit Card Fraud, Part V, Final Models

  I complete the hyper-parameter optimisations for the random forest and xgboost models. I then create a final model using these values to produce AUCs of 0.852 and 0.872.

  May 30, 2020