Program

The 8th edition of the Junior Conference on Data Science and Engineering (JDSE 2023) will take place at the Laboratoire de Mathématiques d’Orsay – Université Paris-Saclay.
The conference is scheduled to be held from September 27th to September 28th.

08:45 – 09:15 Coffee & Registration
09:15 – 09:30 Opening speech
09:30 – 10:30 Keynote: Artificial Intelligence Lab of SLB: How to Use Data Science for Energy Applications

Speaker: Josselin Kherroubi, Director of AI lab, SLB

11:00-11:20 Simulation of Ancient DNA Sequences Using Transformer-based Techniques
11:20-11:40 Auto-Survey Challenge: Advancing the Frontiers of Automated Literature Review
11:40-12:00 Feature Exploration and Augmented Explainability using Self Reinforcement Attention
12:00-12:20 Toward Job Recommendation for All

13:30 – 14:30 Keynote: Less Prep: Data Science between Databases and Machine Learning

Speaker: Gaël Varoquaux, Research director, INRIA

How to Effectively Train an Ensemble of Faster R-CNN Object Detectors to Quantify Uncertainty
From Convolutional Sparse Coding to *-NMF Factorization of Time-Frequency Coefficients
A Comprehensive Benchmarking Study of Self-Supervised and Weakly Supervised Learning on Histology Datasets
Towards Creating Longer Genetic Sequences with GANs: Generation in Principal Component Space
Investigating Transformers for Human Action Forecasting
Document AI: OCR Free Pre-trained Transformer Models for Energy Industry
Towards Anonymized Speech Synthesis for a Low-Resource Language
A Toolchain to Ease the Evaluation of Abstract Automotive E/E Architectures
The Challenging Neural Decoding with General-Purpose Networks and its Improvement via Probabilistic Embeddings
Insight into Fair Universe: Binary Classification of Tabular Data Affected by Systematic Uncertainty
Multi-Model Performance Comparison and Selection on Different CAN Intrusion Datasets
Enhancing Vehicles’ Trajectory Prediction using Occupancy Grids and Temporal Convolutional Networks
Mining Bias-Target Alignment from Voronoi Cells

Speaker: Jean Ponce, Professor, Ecole Normale Supérieure-PSL and New York University

11:00-11:20 Equivariant Single-Stage Object Detection
11:20-11:40 CompoNeRF: Visualizing scenes with Objects of Interest by Composing Neural Radiance Fields
11:40-12:00 Entropy Guided Pruning Algorithm for Reducing the Size of Deep Neural Networks
12:00-12:20 The Cure for Sparse Double Descent