Hassan Maissoro

Greater Paris Metropolitan Region

I hold a Diplome d’Ingenieur en Data Science & Genie Statistique and a PhD in Statistics, both from CREST-ENSAI. During my PhD, I worked at the intersection of nonparametric statistics, functional data analysis, and time series, with applications in energy through a CIFRE convention with DataStorm, a data science consultancy. I also contributed at Capgemini Invent Lab to the open-source Python package MAPIE, a scikit-learn-compatible library for prediction intervals, uncertainty quantification, and risk control in machine learning. I am now actively looking for industry positions starting May 2026 in data science, quantitative research, or applied statistics.

Research Interests

My work spans both methodological and applied directions:

Functional Data Analysis (FDA): nonparametric estimation for curves and trajectories
Time Series & Forecasting: dependent functional data, adaptive prediction
Statistical Learning: regularization, model selection, high-dimensional inference
Machine Learning: supervised/unsupervised methods, neural networks for structured data
Quantitative Finance: log-return forecasting
Energy & Environment: load forecasting, anomaly detection in sensor data
Health & Sports Analytics: longitudinal data, performance modeling
Numerical Statistics: efficient implementation of statistical algorithms in R and Python

PhD in a Nutshell

My thesis, carried out at CREST and the data science consultancy DataStorm, focused on adaptive estimation for weakly dependent functional time series with irregular sample paths, under the supervision of Valentin Patilea and Myriam Vimond.

Beyond the technical contributions, the PhD experience sharpened a set of skills directly relevant to industry:

Problem solving: translating ambiguous, open-ended problems into tractable mathematical formulations
Rigorous thinking: designing and proving theoretical guarantees under realistic assumptions
Data-driven validation: extensive simulation studies and real-data experiments to stress-test methods
Scientific communication: writing for top-tier journals and presenting at international conferences
Software development: building robust, well-tested statistical software in R and Python
Project autonomy: driving a multi-year research agenda with shifting constraints and deadlines
Collaboration: working within an academic-industry partnership (CIFRE thesis with DataStorm)

News

Apr 22, 2026	I am open to new opportunities starting May 4, 2026. Feel free to reach out!
Jan 22, 2026	A journal submission for Adaptive Prediction for Functional Time Series is forthcoming, targeting Fall 2026.
Jun 04, 2025	Our paper, Adaptive Estimation for Weakly Dependent Functional Times Series, has been published in the Journal of Time Series Analysis (June 2025).

Selected publications

Adaptive estimation for Weakly Dependent Functional Times Series

Hassan Maissoro, Valentin Patilea, and Myriam Vimond

Journal of Time Series Analysis, 2025

Abs HTML PDF Poster Slides

We propose adaptive mean and autocovariance function estimators for stationary functional time series under \mathbbL^p-m-approximability assumptions. These estimators are designed to adapt to the regularity of the curves and to accommodate both sparse and dense data designs. The sample paths are observed with error at possibly random design points. Data-driven local bandwidths are selected by minimizing explicit quadratic risk bounds that exploit the local regularity of the process. As a first step, we introduce a local regularity estimator and derive a nonasymptotic concentration bound for it. We also derive the asymptotic normality of the mean estimator, which allows honest inference for irregular mean functions. Simulations and a real data application illustrate the performance of the new estimators.
Adaptive prediction for Functional Times Series

Hassan Maissoro, Valentin Patilea, and Myriam Vimond

CREST Working Paper (journal submission forthcoming), 2025

Abs PDF Slides

We propose an adaptive curve prediction method for stationary functional time series (FTS) with irregular sample paths that are observed with error at discrete domain points. Our approach is based on the best linear unbiased predictor and requires knowledge of the mean, covariance and autocovariance functions of the FTS, as well as the conditional variance of the measurement errors. We introduce new adaptive non-parametric estimators for the covariance and autocovariance functions, using adaptive local bandwidth rules. These estimators adapt to the local regularity of the FTS, resulting in improved prediction accuracy. We derive pointwise and uniform convergence rates for the covariance and autocovariance estimators. We also establish a uniform convergence rate for the adaptive mean estimator. A simulation study and a real data application illustrate the good performance of the new predictor.