I am a Machine Learning Engineer working for HomeChoice International.
Here is a summary of my skillset, and tools that I am proficient with:
(last update 2022-01-21)
Tool | Approximate Hours of Usage/Experience |
---|---|
SQL (Vertica, MySQL, AWS Athena, Google BigQuery, PostgreSQL) | 3300 |
R (incl. tidyverse) | 2842 |
Python (incl. pandas, numpy, scipy, sklearn, datetime) | 1360 |
Microsoft Excel | 1100 |
Supervised Learning & general Machine-Learning theory (transfer learning, classication/regression: neural nets, GBM family, random forest, tree-based models, linear models, elastic-net/penalised linear models, regression splines, local linear models, KNN, bias/variance tradeoff, cross-validation etc.) | 621 |
ML-Ops (models in production) | 532 |
Theory of Experimental Design (null hypothesis testing, bayesian inference, orthogonal designs, power calculation, unbiased variance reduction techniques …) | 501 |
Recommender Engines (DCN, Wide&Deep, Factorisation Machines, Collaborative Filtering incl. Matrix Factorisation, Content-Based, cold-start recommenders) | 376 |
TensorFlow + Keras | 169 |
RMarkdown & LaTeX | 161 |
Uplift Modelling (model-based estimation of heterogeneous treatment effects) | 152 |
Image-based Models (segmentation & masking, classification, multi-class classification (auto tagging), tensorflow/keras, transfer learning | 150 |
Association Rule Mining | 100 |
Exponential Smoothing Models (time series prediction) | 70 |
Natural Language Processing (NLP): Word & Document Embedding Methods | 70 |
Multi-Armed Contextual Bandit Algorithms | 65 |
Meta-Heuristic Optimisation (genetic algorithms, simulated annealing, TABU search) | 60 |
Web Scraping (BeautifulSoup & Selenium in Python) | 53 |
Clustering algorithms (k-means, k-medoids, hierarchical family, CLARA, DBSCAN) | 50 |
Financial Modelling | 50 |
ARIMA, SARIMA, ARIMAX models (time series prediction) | 50 |
H2o Machine Learning Framework (R library) | 40 |
Non-Linear Dimension Reduction (T-SNE, UMAP, ISOMAP, Locally-Linear Embedding) | 36 |
R Shiny | 30 |
Multi-Variable Analysis (PCA, Factor Analysis, SVD Bi-Plots, Canonical Correlation Analysis) | 30 |
Causal Inference: Bayesian Networks and Do-Calculus | 25 |
HTML, CSS and JavaScript | 20 |
Git | 16 |
Linear Optimisation | 15 |
State Space Models (time series prediction) | 12 |
General Reinforcement Learning Theory | 10 |
Multi-Objective Optimisation | 10 |
Copulas (Financial Modelling) | 5 |
Arch/Garch Models (time series variance prediction/inference) | 2 |
TBATS (time series prediction) | 2 |
…this list tbc | 0 |
Website built using R-markdown
Website hosted by GitHub
Image to ASCII-art conversion using https://cloudapps.herokuapp.com/imagetoascii/
Piano
Jazz
Origami
Calisthenics
Rubik’s Cube
Football
Other hobbies, which I have accepted that I will only have time for again after I retire: