Infor

Intern – Comparative Analysis of ML and Time Series-Based Forecasting Engines

4/5-month duration from February 2024

Mission/Challenges:

The intern will conduct a comprehensive comparison between the AI Engine (ML-based) and the existing Time Series (TS) engine, which are in separate code bases (Python and Java, respectively). The challenge is to devise an intelligent methodology for comparing these engines, considering error metrics and the potential use of holdout mode. The goal is to come up with smart rules to determine which engine to use for each item/location, while considering cost implications, such as increased AWS costs due to holdout mode. The intern will work on creating a controller for engine selection.

Skills and Requirements:

• Proficiency in both Python and Java.
• Understanding of machine learning and time series forecasting principles.
• Strong data analysis skills.
• Knowledge of error metrics and statistical analysis.
• Cost-awareness for cloud computing.
• Good communication skills for presenting findings and recommendations.