Kouyakou-Abalo SIMSOBA

Kouyakou-Abalo SIMSOBA

Google DeepMind Scholar | M.Sc. Mathematics (PAUSTI, African Union Scholar) | MSc in AI for Science at Stellenbosch University

Curriculum Vitae

Download Full CV (PDF)

Research Interests

  • Robust Machine Learning: Methodologies and loss functions resilient to noisy data and outliers
  • Time Series Forecasting: Deep learning approaches for temporal prediction
  • NLP & LLMs: Natural Language Processing and Large Language Model applications
  • Explainable AI (XAI): Model interpretability and transparency
  • Computer Vision & Speech Processing: Advanced image and signal analysis

Education

  • M.Sc. in Artificial Intelligence for Science (2025 – Present)
    African Institute for Mathematical Sciences (AIMS), South Africa
    Google DeepMind Scholar

  • Joint M.Sc. in Mathematics (Statistics Option) (2023 – 2025)
    Pan African University Institute for Basic Sciences, Technology and Innovation (PAUSTI), Kenya
    African Union Scholar
    Thesis: A Hybrid Minkowski–Log–Cosh loss function for robust LSTM-based time series forecasting

  • B.Sc. in Mathematics, Statistics and Socio-economic Applications (2018 – 2021)
    Université de Kara, Togo
    Thesis: Evaluating the effects of online learning on the student population of the Université de Kara during COVID-19

Peer-Reviewed Publications

A Hybrid Minkowski-Log-Cosh Loss Function for Robust LSTM-Based Time Series Forecasting

Published in IEEE Access, 2025

This research introduces a novel Hybrid Minkowski–Log–Cosh (MLC) loss function for robust LSTM-based time series forecasting in noisy environments. The Log-cosh loss function provides a tunable framework where parameter $p$ controls the trade-off between outlier robustness and error sensitivity. Validated on 11 years of malaria incidence data, MLC achieves 18.23% MAPE (vs. 20.96% for MSE) in clean data regimes and reduces mispredicted cases by approximately 117,000 in outlier-contaminated scenarios. The method maintains $\mathcal{O}(N)$ computational complexity while offering explicit gradient analysis and plug-and-play compatibility with deep learning models.

Recommended citation: Simsoba, K.-A., Oscar, N., & Mageto, T. (2025). "A Hybrid Minkowski-Log-Cosh Loss Function for Robust LSTM-Based Time Series Forecasting." IEEE Access, 13, 187307–187319.
Download Paper

Research Experience

  • Graduate Research Project, Pan African University (PAUSTI) (2024–2025)
    • Developed a hybrid Minkowski–Log–Cosh loss function to improve robustness of LSTM-based forecasting
    • Applied models to malaria incidence prediction using 10 years of epidemiological data
  • Undergraduate Research Project, Université de Kara (2021–2022)
    • Conducted statistical analysis of online learning outcomes during the COVID-19 pandemic

Teaching Experience

  • Mathematics and Physics Teacher, Public High Schools, Kara, Togo (2019 – 2023)
    Delivered secondary-level instruction emphasizing conceptual understanding, analytical reasoning, and systematic problem-solving.

    Mathematics Modules

    • Probability and Combinatorics
    • Set Theory and Logic
    • Spatial Geometry (3D)
    • Polynomial and Homographic Functions
    • Continuity and Differentiability
    • Arithmetic and Geometric Sequences
    • Complex Numbers

    Physics Modules

    • Kinematics and Motion under Gravity
    • Mechanical and Electrical Oscillations
    • Wave Phenomena and Propagation
    • Electrostatics and RLC Circuits
    • Thermodynamics

  • Tutor (Volunteer), École Polytechnique de Lomé, Togo (2023)
    Graduate tutoring for Master’s students in the Big Data program.

    Inferential Statistics

    • Hypothesis Testing
    • Confidence Intervals
    • Likelihood-Based Inference
    • Sampling Distributions and Asymptotic Results

    Optimization

    • Unconstrained and Constrained Optimization
    • Lagrange Multipliers and KKT Conditions
    • Gradient-Based Methods
    • Convex Optimization

Technical Skills

  • Programming Languages: Python (Advanced), R, MATLAB, SQL, C/C++
  • AI / ML Frameworks: TensorFlow, PyTorch, Scikit-learn, Keras
  • Tools: Power BI, STATA, GIS, LaTeX, GitHub, Jupyter, Overleaf

Awards & Honors

  • Google DeepMind Scholarship, AIMS South Africa (2025 – 2026)
  • African Union Scholarship, PAUSTI Kenya (2023 – 2025)
  • Togolese National Government Scholarship (2015 – 2018)