CAIN 2024
Sun 14 - Mon 15 April 2024 Lisbon, Portugal
co-located with ICSE 2024

As the integration of Machine Learning (ML) models becomes pervasive in software systems, the associated energy costs have emerged as a critical concern. This research delves into the energy efficiency of ML-enabled systems, focusing on the ML component itself and its impact on the overall system. Our primary emphasis is on the data-centric approach, particularly in the context of feature selection and handling concept drift, and how these energy-efficient components affect the overall energy consumption of ML-enabled systems. In our initial investigation, we explored feature selection methods and identified significant variations in their energy consumption. This led us to delve deeper into understanding how different techniques for scoring features contribute to the overall energy footprint of ML models. Subsequently, we examined the impact of changes in data distribution, often referred to as concept drift, on model accuracy and the associated energy costs. Our empirical experiments revealed noteworthy insights into energy-efficient strategies for handling concept drift, a crucial aspect of maintaining ML-enabled systems. We compared various methods and their effectiveness in mitigating the adverse effects of concept drift while keeping energy consumption in check. The findings from our research contribute to the development of sustainable and energy-efficient ML models within the broader context of software engineering. Lastly, we will compare how different alternatives of ML components in ML-enabled systems affect the overall energy consumption of ML-enabled systems.

Mon 15 Apr

Displayed time zone: Lisbon change

11:00 - 12:30
Doctoral Symposium and Energy-Aware AI EngineeringDoctoral Symposium / Research and Experience Papers at Pequeno Auditório
Chair(s): Justus Bogner Vrije Universiteit Amsterdam, Silverio Martínez-Fernández UPC-BarcelonaTech
11:00
6m
Talk
Software Design Decisions for Greener Machine Learning-based Systems
Doctoral Symposium
Santiago del Rey Universitat Politècnica de Catalunya (UPC)
11:06
6m
Talk
Energy-Efficient Development of ML-Enabled Systems: A Data-Centric Approach
Doctoral Symposium
11:12
6m
Talk
Optimizing Data Analytics Workflows through User-driven Experimentation
Doctoral Symposium
Keerthiga Rajenthiram Vrije Universiteit Amsterdam
11:18
6m
Talk
Component-based Approach to Software Engineering of Machine Learning-enabled Systems
Doctoral Symposium
Vladislav Indykov Chalmers | University of Gothenburg
11:24
6m
Talk
Threat Modeling of ML-intensive Systems: Research Proposal
Doctoral Symposium
Felix Viktor Jedrzejewski Blekinge Institute of Technology
11:30
6m
Talk
Continuous Quality Assurance ML Pipelines under the AI Act
Doctoral Symposium
Matthias Wagner Lund University
11:36
10m
Talk
Green Runner: A tool for efficient deep learning component selection
Research and Experience Papers
Jai Kannan Applied Artificial Intelligence Institute, Deakin University, Scott Barnett Applied Artificial Intelligence Institute, Deakin University, Anj Simmons , Taylan Selvi Applied Artificial Intelligence Institute, Deakin University, Luís Cruz Delft University of Technology
11:46
15m
Talk
Engineering Carbon Emission-aware Machine Learning Pipelines
Research and Experience Papers
Erik Johannes Husom SINTEF Digital, Sagar Sen , Arda Goknil SINTEF Digital
12:01
10m
Talk
Identifying architectural design decisions for achieving green ML serving
Research and Experience Papers
Francisco Durán Universitat Politècnica De Catalunya - Barcelona Tech, Silverio Martínez-Fernández UPC-BarcelonaTech, Matias Martinez Universitat Politècnica de Catalunya (UPC), Patricia Lago Vrije Universiteit Amsterdam
Pre-print
12:11
10m
Talk
Green AI: a Preliminary Empirical Study on Energy Consumption in DL Models Across Different Runtime Infrastructures
Research and Experience Papers
Negar Alizadeh Universiteit Utrecht, Fernando Castor University of Twente and Federal University of Pernambuco
12:21
9m
Live Q&A
Energy: Q&A Session
Research and Experience Papers