Multi-task MLP
HoldTechniques
A multi-layer perceptron trained to predict multiple related outputs at once.
Why it's here
Placed in Hold: 1 article(s) of evidence from 1 source(s), led by research-stage coverage, with 0 in the last 30 days. Confidence 24%. Low accumulated evidence, so it defaults conservatively pending more signal.
Evidence (1)
- 5InfoQ·9/2/2026researchSwiggy Builds Multi-Task MLP for Customer Lifetime Value Prediction
Swiggy developed an in-house predicted lifetime value model using more than 350 pre-order features for Food and Instamart. The company used a multi-task MLP with order count as an auxiliary task, cutting model parameters by 63% while improving predictive performance, and then applied the pLTV signal to Google Target ROAS bidding for customer acquisition.