BestPromptFinder

Machine Learning Project Plan From Problem to Model

Quality, usefulness and confidence are AI evaluations out of 100 — not user ratings. Community results come from visitors' "worked / didn't work" votes below. How scoring works →

The prompt

Help me plan a machine learning project properly before I start training models. Business problem: [e.g. predict which customers will cancel, forecast weekly demand, classify support tickets] Data available: [sources, rows, columns, labels, time period] How the prediction will be used: [who acts on it and how] Constraints: [latency, explainability, privacy] Give me: 1. The problem framed as an ML task (classification, regression, forecasting) with the target defined precisely. 2. A baseline to beat, starting with a simple rule or model. 3. Data checks: leakage risks, class imbalance, missing values, train/validation/test split (time-based if needed). 4. Features to try, and a first model and why. 5. Metrics that match the business cost of mistakes, not just accuracy. 6. An experiment log template and a plan for monitoring after launch. Tell me honestly if the data seems too small or too messy for ML, and what to do instead.

Why this prompt

• Hand-picked, editorially reviewed prompt
• Written for ChatGPT, Claude, Gemini

Source

Created by BestPromptFinder — editorial prompt, not user-generated. See our content & source policy.

Related Data / Analysis prompts