Gemini Optimized

Best Gemini prompts for Models

A specialized toolkit of advanced AI prompts designed specifically for Models.

Professional Context

Balancing the daily priorities of model validation and data preprocessing is a constant struggle, as ensuring the quality of our industry-specific databases while meeting tight time-to-completion deadlines can be overwhelming, especially when error rates are on the line.

💡 Expert Advice & Considerations

Don't waste your time trying to automate everything with Gemini, focus on using it to augment your data interpretation skills and workflows within the Google ecosystem.

Advanced Prompt Library

4 Expert Prompts
1

Model Performance Comparison Across Datasets

Terminal

Given a set of classification models trained on different datasets, compare their performance using metrics such as accuracy, precision, recall, and F1-score, and provide a detailed analysis of the results, including visualizations and statistical significance tests, using Google Data Studio to create interactive dashboards and Google Sheets for data manipulation.

✏️ Customization:Replace the dataset names and model types with your own specific use case.
2

Time Series Forecasting with Seasonal Decomposition

Terminal

Using a dataset of monthly sales data from the past five years, perform seasonal decomposition to identify trends, seasonality, and residuals, and then use a combination of Google AutoML and TensorFlow to train a time series forecasting model, evaluating its performance using metrics such as mean absolute error and mean squared error, and provide a report on the results, including visualizations and recommendations for future improvements.

✏️ Customization:Update the dataset and time series frequency to match your specific problem.
3

Feature Engineering for Clustering Analysis

Terminal

For a given dataset of customer demographic and transactional data, perform feature engineering to extract relevant features, including scaling and encoding categorical variables, and then use Google Colab to train a clustering model, evaluating its performance using metrics such as silhouette score and calinski-harabasz index, and provide a detailed report on the results, including visualizations and insights into customer segments.

✏️ Customization:Replace the dataset and feature engineering techniques with your own specific requirements.
4

Model Interpretability using SHAP Values

Terminal

Using a trained regression model and a dataset of feature importance values, calculate SHAP values to explain the contribution of each feature to the predicted outcomes, and provide a detailed analysis of the results, including visualizations and insights into feature interactions, using Google Data Studio to create interactive dashboards and Google Sheets for data manipulation, and evaluate the results against a set of quality metrics, including feature correlation and partial dependence plots.

✏️ Customization:Update the model type and dataset to match your specific use case and quality metrics.