Everything you need to know about using MCP Analytics with Claude Desktop
Get up and running with MCP Analytics in minutes
Extensive collection of statistical and machine learning tools at your fingertips
Classic regression analysis with diagnostics and predictions
RegressionL2 regularized regression for multicollinearity
RegressionFeature selection with L1 regularization
RegressionCombined L1 and L2 regularization
RegressionUnsupervised clustering for segmentation
Machine LearningEnsemble learning with decision trees
Machine LearningGradient boosting for high performance
Machine LearningBinary and multinomial classification
Machine LearningDimensionality reduction and visualization
Machine LearningEstimate treatment effects from observational data
Causal InferenceCreate counterfactuals for policy evaluation
Causal InferenceQuasi-experimental design for causal effects
Causal InferenceBayesian time series causal analysis
Causal InferenceTime series forecasting with seasonality
Time SeriesFacebook's forecasting tool for business data
Time SeriesMultivariate time series analysis
Time SeriesCompare means between two groups
Hypothesis TestingCompare means across multiple groups
Hypothesis TestingTest independence of categorical variables
Hypothesis TestingProbabilistic approach to experiment analysis
Hypothesis TestingCustomer segmentation by behavior
Customer AnalyticsPredict customer lifetime value
Customer AnalyticsExplore relationships between variables
ExploratorySurvival analysis and hazard modeling
SurvivalMeasure demand sensitivity to price changes
EconomicsMarket basket analysis with Apriori algorithm
AssociationExample conversation with Claude
You: "I have sales data with advertising spend and revenue. Can you help me understand the relationship?" Claude: I'll help you analyze the relationship between advertising spend and revenue using linear regression. Let me run this analysis for you. [Claude runs the analysis using MCP Analytics] Claude: Here's what I found: - Strong positive correlation (r=0.84) between ad spend and revenue - Each $1,000 in advertising generates approximately $4,200 in revenue - The model explains 71% of revenue variation (R² = 0.71) - View the full interactive report: [link] The analysis shows statistical significance (p < 0.001) and includes: - Scatter plot with fitted line - Residual diagnostics - Confidence intervals - Predictive capabilities Would you like to explore specific segments or add more variables?
Connect Claude Desktop to MCP Analytics and unlock powerful statistical capabilities