P
📊Analyse de donnéesIntermediateAll AIs

Select features using multiple methods

Select the most predictive features

Paste in your AI

Adapt this prompt

Paste this prompt in ChatGPT, Claude or Gemini and customize the variables in brackets.

Write Python code to perform feature selection on DataFrame df for predicting [target] using: filter methods (mutual information, chi-squared), wrapper method (recursive feature elimination with cross-validation), and embedded method (Lasso regularization). Compare the selected feature sets and recommend a final set.

Personalize this prompt with Léa · AI assistant

Léa rewrites this prompt for your job and your exact goal — 3 quick questions.

Use Cases

Select the most predictive features

Improve this prompt

Run this prompt through the Optimizer to strengthen its context, constraints and expected format.

Improve this prompt with the Optimizer

Comments

  • LéaAI

    Pensez à standardiser les variables avant le chi² et le Lasso, et adaptez la mesure d’information mutuelle au type de cible (classification vs régression). Croisez les résultats sur plusieurs splits pour vérifier la stabilité des sous-ensembles retenus, et privilégiez l’intersection des méthodes si la performance est équivalente.

📬 Get new prompts every week

Join our newsletter and never miss a prompt.

Go further

Similar Prompts

📊Analyse de donnéesIntermediateGemini

Gemini Prompt for Extracting Data Insights

Extracting insights from raw data is one of the major challenges faced by analysts, data scientists, and decision-makers. Gemini, thanks to its multimodal reasoning capability and extended context window, excels at analyzing large datasets to extract meaningful patterns, anomalies, and actionable recommendations. Unlike traditional BI tools that require predefined queries, Gemini can explore your data in an exploratory manner, identify non-obvious correlations, and formulate hypotheses you might not have considered. Whether you work with CSV files, database exports, financial reports, or application logs, a well-structured prompt allows Gemini to transform columns of numbers into strategic narratives. This guide provides you with optimized prompts to get the most out of Gemini in your data analyses, covering both descriptive and predictive analysis, with variants adapted to your level of expertise and the complexity of your datasets.

483
📊Analyse de donnéesIntermediateAll AIs

Perplexity Prompt for Analyzing User Feedback

Analyzing user feedback is a strategic lever for improving a product, service, or customer experience. Yet, faced with hundreds or even thousands of scattered reviews—Google reviews, social media comments, support tickets, and NPS surveys—the task quickly becomes time-consuming. Perplexity, with its augmented search and intelligent synthesis capabilities, can turn this mass of qualitative data into actionable insights. Unlike manual analysis that would take hours, Perplexity can cross-reference sources, identify recurring trends, and categorize feedback by theme in minutes. Whether you're a product manager prioritizing your roadmap, a CX manager looking to reduce churn, or a startup founder validating a pivot, a well-structured prompt transforms Perplexity into a true quality analyst. This guide offers an optimized prompt to extract maximum value from your user feedback, with variants adapted to your expertise level and need complexity.

338
📊Analyse de donnéesIntermediateAll AIs

GitHub Copilot Prompt for Data Analysis

GitHub Copilot, the AI assistant developed by GitHub and OpenAI, is not limited to code generation. It is a powerful tool for analyzing data directly in your development environment. Whether you work with CSV files, SQL databases, or pandas DataFrames, Copilot can help you explore, clean, transform, and visualize your data without leaving your IDE. By formulating precise prompts, you can ask it to detect anomalies, compute descriptive statistics, identify correlations, or generate relevant graphs. The major advantage of Copilot for data analysis lies in its ability to understand the context of your existing code: it adapts to the libraries you use (pandas, numpy, matplotlib, seaborn) and offers analyses consistent with your data structure. This guide presents the best prompts to leverage GitHub Copilot in your data analysis tasks, from initial cleaning to creating actionable visual reports.

437

Initial Exploratory Analysis

Start a structured EDA

424