Tatenda MawangoData scientist and PhD researcher

I'm Tatenda, a data scientist doing a PhD in machine learning.

I started at Wrocław University of Science and Technology in October 2026, after a BSc in informatics and an MSc in data science and AI at Kaunas University of Technology. For my MSc thesis I forecast restaurant demand from about 12.9 million POS records, then checked the prediction intervals. The 80% ones covered about 57%.

Read the thesis write-up

Drag across the chart or use the arrow keys to read any month.

Nov 2026

PhD in machine learning, Wrocław University of Science and Technology

Month 2 of about 41, 5% of the way.

My education since 2018. The PhD part is a forecast: I expect to finish around February 2030, and the 50, 80 and 96% bands are illustrative, not a fitted model.

The thesis: do the intervals mean what they say?

My MSc thesis: order forecasts for 17 Vilnius restaurants with ARIMA, DeepAR and a TFT, then a test of whether their prediction intervals hold.

Venues
17
POS records
~12.9M
Models compared
4
TFT daily MAE
37.3
Total80% interval coverage
57%

Read the full write-upCode and thesis PDF on GitHub

  1. 50% interval28.6% of 50%
  2. 80% interval57.1% of 80%
  3. 96% interval75.7% of 96%
What the Temporal Fusion Transformer's (TFT) daily intervals promised against what they delivered, over 14 samples and five forecast days. Every interval is too narrow: the 80% band contained the actual value only 57.1% of the time. Values read from the thesis calibration plot.

More work, from labs to systems

All 23 projects
A random walk with drift forecast of the gold price: the intervals widen the further ahead it looks.

Statistical Analysis & Forecasting

2024Descriptive and predictive analytics in R: PCA, clustering, ARIMA and ETS forecasting and multivariate modeling, all reported through RMarkdown.

Covers
PCA, k-means, ARIMA, ETS
Reports
RMarkdown
  1. 2024Mathematical Methods of AIA grab-bag of mathematically grounded AI work: explainable AI, recommender systems, Kalman filters and LDA topic modeling, with a focus on interpretability.DataPython, NumPy, scikit-learn
  2. 2024ML Lab ProjectsSupervised learning labs across regression, classification (LDA, QDA, Logistic), KNN, and model evaluation.DataPython, scikit-learn, pandas
  3. 2024AI Lab Methods & ProjectsLab-based AI projects covering ANN classification, Naive Bayes spam detection, time-series forecasting, and statistical data preprocessing.DataPython, Keras, scikit-learn
  4. 2024CUDA Student GradingGPU-parallel student grading in CUDA: filtering and formatting student records by name and grade across thousands of threads.EngineeringCUDA, C++, NVCC
  5. 2024Location-Based Android AppAn indoor positioning Android app that locates the user from Wi-Fi RSSI fingerprints with k-nearest neighbors. Built with Room, Retrofit and Material 3.EngineeringKotlin, Android, Room
  6. 2023Distributed Container SimulationMulti-client distributed simulation in C# using RabbitMQ. Clients adjust container mass under pressure; the system resets on implosion or explosion.EngineeringC#, .NET, RabbitMQ

Work history

  1. Dec 2024 to Mar 2025

    AI & Automation Workflow Specialist

    1man.ai, Kaunas, Lithuania

  2. Aug 2024 to Oct 2024

    IT Support Specialist

    Basic Economy sp. z o.o., Kielce, Poland

  3. Mar 2024 to Aug 2024

    Android Developer Intern

    Basic Economy sp. z o.o., Kielce, Poland

Get in touch

Have a research idea, a role, or a question about one of these projects?

There's also a contact form.