Nonparametric Analysis of US Dairy Production and Consumption

This project was developed for the course of "Nonparametric Statistics" held at Politecnico di Milano in the academic year 2022-2023.

🏆 This was the highest ranking project in its year batch.

We present a nonparametric analysis of the evolution of dairy production and consumption in the United States from 1980 to 2021. Starting from the needs of a stakeholders which is about to enter the market, we present a step-by-step analysis and the findings to provide meaningful information to optimize production and pricing. We also provide an outlier analysis in terms of years and, in the last part, a spatial analysis and a clustering of the types of cheese.

Table of contents

• Installation
  • How to clone the repository
  • How to install the packages
• Analysis and results
• Authors

Installation

How to clone the repository

git clone https://github.com/SmearyTundra/nonparametric-analysis-US-dairy-production-consumption

How to install the packages

Install the required packages from CRAN

packages_list <-
    c(
        "tidyverse",
        "ggplot2",
        "mgcv",
        "rgl",
        "splines",
        "conformalInference",
        "pbapply",
        "parallel",
        "DepthProc",
        "progress",
        "dbscan",
        "beadplexr",
        "robustbase",
        "readxl",
        "tidyr",
        "car",
        "sp",
        "visreg",
        "mgcViz",
        "usmap",
        "raster",
        "sf",
        "maps",
        "ggspatial",
        "BNPTSclust",
        "roahd",
        "fda.usc",
        "npsp"
    )
install.packages(packages_list)

Analysis and results

The repository contains different files to perform the analysis, here we report their explanation together with a hyperlink to the knitted PDF version:

• 01-Conformal-Prediction.pdf contains the implementation of the prediction intervals using a conformal approach.
• 02-Permutation-Tests-for-GAM.pdf contains the permutation tests performed to reduce GAM and keep significant covariates only.
• 03-GAM.pdf provides the implementation for the GAM model and the reverse percentile bootstrap confindence intervals.
• 04-Robustness.pdf contains the robust regression used to detect years outliers.
• 05-Spatial-GAM.pdf empowers a GAM using spatial coordinates, while 05-Spatial-Nonparametric.pdf implements a nonparametric kriging.
• 06-Functional-Depth.pdf contains the Bayesian nonparametric clustering as well as an exploratory analysis of such clusters using depth measures.

The final presentations can be found here:

• 01 Midterm
• 02 Endterm
• 03 Final

The final report can be found here:

• Nonparametric_Analysis_of_US_Dairy_Production_and_Consumption.pdf

Authors

• Teo Bucci (@teobucci)
• Filippo Cipriani (@SmearyTundra)
• Gabriele Corbo (@gabrielecorbo)
• Andrea Puricelli (@apuri99)