{ "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": { "slideshow": { "slide_type": "skip" }, "tags": [ "hide-input" ] }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\n" ], "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "%%html\n", "\n", "\n", "\n", "" ] }, { "cell_type": "markdown", "metadata": { "tags": [ "remove-cell" ] }, "source": [ "---\n", "license:\n", " code: MIT\n", " content: CC-BY-4.0\n", "github: https://github.com/ocademy-ai/machine-learning\n", "venue: By Ocademy\n", "open_access: true\n", "bibliography:\n", " - https://raw.githubusercontent.com/ocademy-ai/machine-learning/main/open-machine-learning-jupyter-book/references.bib\n", "---" ] }, { "cell_type": "markdown", "metadata": { "colab_type": "text", "id": "0MRC0e0KhQ0S", "slideshow": { "slide_type": "slide" } }, "source": [ "# Model Selection" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Introduction\n", "\n", "* Model Selection is the process of choosing the best model among all the potential candidate models for a given problem. \n", "* The aim of the model selection process is to select a machine learning algorithm that evaluates to perform well against all the different parameters." ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Outline\n", "\n", "* Over-fitting and under-fitting\n", "* Bias variance tradeoff\n", "* L1 and L2 Regularization\n", "* Early stopping\n", "* Dropout\n", "* Tuning the hyper-parameters of an estimator" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Regression\n", " \n", "![](../images/model-selection/under_over_justalright.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Regression\n", " - Training data points\n", "\n", "![](../images/model-selection/bias-variance-datapoints.jpg)\n" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Regression\n", " - Over-fitting model fits very well on training data\n", "\n", "![](../images/model-selection/bias-variance-overfitting.jpg)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Regression\n", " - Over-fitting model fits poorly on test data\n", "\n", "![](../images/model-selection/bias-variance-overfitting-testdata.jpg)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Regression\n", " - Under-fitting model fits poorly on training data\n", "\n", "![](../images/model-selection/bias-variance-underfitting.jpg)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Regression\n", " - Under-fitting model fits poorly on test data\n", "\n", "![](../images/model-selection/bias-variance-underfitting-test-data.jpg)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Over-fitting and under-fitting\n", "\n", "- Classification\n", "\n", "![](../images/model-selection/classification.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Bias variance tradeoff\n", "\n", "- Graphical illustration of variance and bias \n", "\n", "![](../images/model-selection/graphicalillustration.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Bias variance tradeoff\n", "\n", "- Model complexity v.s. error\n", "\n", "![](../images/model-selection/total_error.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## L1 and L2 regularization\n", "\n", "$$L2\\ Loss = Loss + {\\lambda}\\sum_{i} w_i^2$$\n", "\n", "$$L1\\ Loss = Loss + {\\lambda}\\sum_{i} \\lvert w \\rvert$$" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## L1 and L2 Regularization\n", "\n", "![](../images/model-selection/circlesquare.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## L1 and L2 Regularization\n", "\n", "![](../images/model-selection/L1L2contour.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## L1 and L2 Regularization\n", "\n", "![](../images/model-selection/ridgelassoItayEvron.gif)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## L1 and L2 Regularization\n", "\n", "![](../images/model-selection/p-norm_balls.webp)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## L1 and L2 Regularization\n", "- The impact of the value of $\\lambda$ \n", " \n", "![](../images/model-selection/lagrange-animation.gif)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Early stopping\n", " \n", "![](../images/model-selection/traintestoverfitting.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Dropout\n", " \n", "![](../images/model-selection/dropoutgif.gif)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Prediction after dropout\n", " \n", "![](../images/model-selection/kUc8r.jpg)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "slide" } }, "source": [ "## Conclusions\n", "\n", "- Training size matters\n", "\n", "![](../images/model-selection/ZahidHasan.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Conclusions\n", "\n", "- How to choose a good model\n", "\n", "![](../images/model-selection/steps.png)" ] }, { "cell_type": "markdown", "metadata": { "slideshow": { "slide_type": "subslide" } }, "source": [ "## Conclusions\n", "\n", "![](../images/model-selection/Bias-vs.webp)" ] } ], "metadata": { "celltoolbar": "Slideshow", "colab": { "authorship_tag": "ABX9TyOsvB/iqEjYj3VN6C/JbvkE", "collapsed_sections": [], "machine_shape": "hm", "name": "logistic_regression.ipynb", "provenance": [], "toc_visible": true }, "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.9" } }, "nbformat": 4, "nbformat_minor": 1 }