← 02 Data Exploration
08 · AI Target YX Optimization
AI Target YX Optimization
Auto-selects the best algorithm for Y and picks improvement-target groups → supports What-if and Reverse What-if (auto-recommends optimal X combinations for a target Y; predicts Y change from X change).
60
Competing algorithms
Bottom 20%
Improvement-target group
Top 20%
Benchmark group
Capabilities
Capabilities
Best-Algorithm Selection
Auto-selects the best algorithm for the Y variable from ~60 competing algorithms.
Target Group Selection
Derives increase/decrease plans for specific records (city, district, store, factory, person, machine) by overall average, bottom 20%, top 20%.
What-if Analysis
Predicts Y from X change; simulates Y when controllable X variables are adjusted.
Reverse What-if Analysis
Sets a target Y and reverse-solves the optimal X combination, with an AI summary (GPT report).
Result
Risk and performance optimization across cities, facilities, sales, failures, and disasters
PPTX Evidence
Source Evidence
Source — XimTier menu PPTX · Slide 06 · 31
