{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "## Nonparametric Models: kNN, Decision Trees and Local Regression" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Author: Omar Al-Ghattas" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Introduction\n", "In this lab we will look at nonparametric modelling, which is an alternative to parametric modelling that has been our focus in the last few weeks. We'll work through kNN, Linear Smoothing and Decision Trees in detail. We will implement these algorithms from scratch, and we'll also explore existing implementations in `sklearn`. We will also work with `Pandas` a bit more heavily to see how it can be used for data analysis." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Nonparametric vs. Parametric Models\n", "Surprisingly, it is difficult to have a good definition for nonparametric models and there is often disagreement about which models are nonparametric rather than parametric. Here we will try to build intuition for the differences between the two types of modelling regimes. First, let's recall our data generating process (DGP) set-up from tutorials, we assume that the data are generated independently according to the following process:\n", "\n", "$$\n", "y = f(x) + \\epsilon,\n", "$$\n", "\n", "where $\\epsilon$ is some random noise, and $f$ is the true data function that we wish to estimate. \n", "\n", "#### Parametric Modelling\n", "In parametric modelling, we place a strong assumption on what kind of function $f$ really is, and we assume that this structure is known when we do our estimation. For example, in linear regression, we assume that $f(x) = w^T x$, i.e. that $f$ is a linear function parameterized by a vector $w$. We then proceed by estimating $w$ from the data. Parametric models tend to have a relatively small number of parameters. They are also usually high bias because of the strong assumptions placed on the underlying model. \n", "\n", "#### Nonparametric Modelling\n", "In contrast to parametric models, here we place little to no assumptions on the underlying function $f$. Nonparametric models (the number of parameters) often grow as the size of the observed data grows.\n", "\n", "#### Regression \\& Classification\n", "There are Nonparametric versions of both regression and classification models, and we will explore many of them throughout the course. Here is a short list to keep in mind:\n", "\n", "Parametric:\n", "1. Linear Regression\n", "2. Ridge Regression\n", "3. Lasso Regression\n", "4. Perceptron (classification)\n", "5. Logistic Regression (both regression and classification)\n", "\n", "Nonparametric:\n", "1. Decision Trees (both regression and classification)\n", "2. k-Nearest Neighbours (kNN) (both regression and classification)\n", "3. Local Linear Regression\n", "4. Support Vector Machines (both regression and classification)\n", "5. Random Forests (both regression and classification)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## kNN\n", "We begin with the simplest nonparametric model, kNN, and we will explore both kNN classification and kNN regression. Both classification and regression variants require an understanding of the k-neighbour function: $\\mathcal{N}_k(x)$. The $k$-neighbour function simply returns the $k$ points in the dataset that are nearest to the input $x$. We can define 'nearest' here quite broadly, and it is up to the modeler to come up with a distance function that they would like to use; the most common choice is the Euclidean distance $\\|\\cdot\\|_2$.\n", "\n", "#### Toy Regression Example\n", "Let's assume we have data $X=[-1,0,1,2,3,4], y=[0.5,0.25,1,0.2, 0.8,3]$, and we choose $k=3$. Given an input point $x=0.25$, we have:\n", "\n", "$$\n", "\\mathcal{N}_3(0.25) = \\{ -1,0,1 \\}.\n", "$$\n", "\n", "Now that we have identified the 3-NN of $x$, we need to return a prediction for it. The idea in kNN is just to use an average over the neighbours, so:\n", "\n", "$$\n", "\\hat{y}(0.25) = \\frac{1}{3} [0.5+ 0.25 +1 ] \\approx 0.58\n", "$$\n", "\n", "is our prediction for this input. More rigorously, an average over the $k$-nearest neighbours can be written:\n", "\n", "\\begin{align*}\n", "\\hat{y}(x) \n", "&= \\frac{1}{k} \\sum_{i \\in \\mathcal{N}_k (x)} Y_i \\\\\n", "&= \\sum_{i=1}^n \\frac{1}{k}\\mathbf{1}\\{\\text{$X_i$ is a kNN of $x$} \\} Y_i.\n", "\\end{align*}\n", "\n", "#### Toy Classification Example\n", "The kNN classifier works in the same was as kNN regression, except that instead of averaging, we take a majority vote of the class labels. This works for multi-class classification as well.\n", "\n", "#### Note\n", "The examples we showed here use numerical data, as in $X$ is comprised of numbers. However, kNN works whenever you have a way of measuring distance between objects. For example, if $X$ represented words in the alphabet that you were trying to classify into given classes, you could use kNN with Levenshtein (edit) distance." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In this lab we'll work with some of the same code used in Lab2 for sampling. For regression problems we'll simulate toy data from the model:\n", "\n", "$$\n", "y=f(x) + \\epsilon, \\qquad f(x) = 0.3 \\cos(x) + 0.4 \\ln(10x), \\quad \\epsilon \\sim N(0,\\sigma^2), \\quad \\sigma=0.5.\n", "$$\n", "\n", "The following code loads in the required packages and defines the `f_sampler` function used in Lab2." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "\n", "def f_sampler(f, n=100, sigma=0.5, seed=123): \n", " \n", " np.random.seed(seed)\n", " \n", " # sample points from function f with Gaussian noise (0,sigma**2)\n", " xvals = np.random.uniform(low=1, high=10, size=n)\n", " yvals = f(xvals) + sigma * np.random.normal(0,1,size=n)\n", " \n", " # build dataset D\n", " D = np.zeros(shape=(n, 2))\n", " D[:,0] = xvals; D[:,1] = yvals; \n", " \n", " return D" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", " \n", "#### Exercise: \n", "In this exercise, we will write code to implement kNN regression from scratch. Generate data using the following code:\n", "\n", "```\n", "f = lambda x: 0.3 * np.cos(x) + 0.4 * np.log(10*x)\n", "fsamples = f_sampler(f, 80, sigma=0.2, seed=120)\n", "X = fsamples[:, 0]\n", "y = fsamples[:, 1]\n", "```\n", "\n", "Write a function `kNNRegression(x, X, y, k)` that implements kNN regression (i.e., predicts a value for `x` where `x` is potentially an array of inputs based on data `X`, `y`) using $k$ neighbours and using euclidean distance. Run the algorithm with $k=1,\\dots, 16$ and $p=2$ and plot each fit on a $4 \\times 4$ grid plot, one plot for each $k$. Be sure to plot the original function and the samples on the same plot as well.\n", "\n", "Hint: the function `np.argsort` might be useful here.\n", " " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### kNN with `sklearn`\n", "Of course, `sklearn` has existing implementations of both kNN classification and kNN regression. These can be found by running:\n", "\n", "```\n", "from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor\n", "```\n", "\n", "Their usage follows the same standard `sklearn` approach we have seen in previous labs." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Linear Smoothing\n", "From the previous exercise, you will most likely notice that the fitted regression function using kNN is quite jagged, especially for smaller values of $k$. To understand this behaviour, it helps to look at the specific form of kNN predictions which we wrote down earlier:\n", "\n", "\\begin{align*}\n", "\\hat{y}(x) &= \\sum_{i=1}^n \\frac{1}{k}\\mathbf{1}\\{\\text{$X_i$ is a kNN of $x$} \\} Y_i.\n", "\\end{align*}\n", "\n", "Note that this is of the form: $\\hat{y}(x) = \\sum_{i=1}^n w_i(x) Y_i$, so we can think of kNN as producing a weighted average of the responses $Y_i$, where the weights are chosen based on the given $x$. The 'weighting' scheme in kNN is rather crude; it looks at the closest $k$ elements and cuts out the rest, resulting in a jagged fit. We can improve the kNN regression model by incorporating other weight functions that will induce a smoother fit - this is the idea behind Linear smoothing, and we use special functions called kernels to define the weighting schemes. Specifically, we \n", "first define some common kernel functions:\n", "\n", "1. Box-car Kernel: $K(u) = \\mathbf{1} \\{ |u| \\le 1/2 \\}$\n", "2. Gaussian Kernel: $K(u) = \\frac{1}{\\sqrt{2 \\pi}} \\exp (-u^2/2)$ \n", "3. Epanechnikov Kernel: $K(u) = \\frac{3}{4} (1-u^2) \\mathbf{1} \\{ |u| \\le 1\\}$\n", "\n", "We then use a chosen kernel to construct a linear smoother which is defined by:\n", "\n", "\\begin{align*}\n", "\\hat{y}(x) \n", "&= \\sum_{i=1}^n \\frac{K \\left (\\frac{\\| x - X_i \\|_2}{h} \\right )}{ \\sum_{j=1}^n K \\left (\\frac{\\| x - X_j \\|_2}{h} \\right ) } Y_i.\n", "\\end{align*}\n", "\n", "In words, the $i$-th weight (the weight assigned to the $i$-th response $Y_i$) is the kernel function evaluated at the distance of $X_i$ from the input point $x$, and divided by a normalizing term to ensure that the weight add up to 1. The parameter $h$ is called the bandwidth, and plays a similar role as $k$ in kNN regression. Larger values of $h>0$ means we incorporate more information from points that are further away." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", " \n", "#### Exercise: \n", "Implement the three kernels defined above." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", " \n", "#### Exercise: \n", "Write a function `LinearSmoother(x, X, y, kernel, h)` that implements LinearSmoother using the data from the previous kNN regression exercise. Demonstrate your code by plotting your fitted functions for various bandwidth values and the three kernels defined above. Use the bandwidths `[0.8, 0.9, 1, 3]` for the boxcar kernel, `[0.5, 0.8, 1, 3]` for the Epanechnikov kernel and `[0.4, 0.5, 1, 5]` for the Gaussian kernel. What do you notice happens to the fits as you vary the bandwidth?" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Existing Implementations\n", "Unfortunately, there are no `sklearn` implementations of the linear smoother, however a good implementation does exist in `scipy.statsmodels.nonparametric.kernel_regression`, (kernel regression, Nadaraya-Watson) are two different names commonly used for linear smoothing. You can read more about it here:\n", "\n", "https://www.statsmodels.org/dev/generated/statsmodels.nonparametric.kernel_regression.KernelReg.html" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## The curse of dimensionality\n", "The algorithms we have seen so far (kNN and Linear Smoothing) are quite straight forward to construct, even more so than the parametric models we have dealt with in the past. One problem that arises however is the curse of dimensionality, which is a problem that plagues much of nonparametric modelling. The problem is the following: distances break down in high dimensional spaces.\n", "\n", "What this means is that when we are dealing with high dimensional data (such as MNIST), the distances between objects in high dimensional space are not very informative. So an image with class label '0' will have similar distance to an image with class label '8' as it does to other images in class '0'. Since kNN and LS rely on finding neighbouring points, their performance is affected negatively, to the point that they become useless.\n", "\n", "We can demonstrate this breakdown of distances through the following simulation experiment:" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The set $[0,1]^d$ consists of $d$-dimensional vectors with elements between 0 and 1 and is called the $d$-dimensional unit cube. We can generate $n$ independent samples uniformly distributed on $[0,1]^d$ using the code: `np.random.random(low=0.0, high=1.0, size=(n,d))`.\n", "\n", "The idea will be to sample $n=1000$ points on $[0,1]^d$ for increasing values of $d$. For each $d$, we will compute the pairwise Euclidean distances between the $n$ points and plot a histogram. As the dimension increases, we will see that the distribution of the distances becomes more and more concentrated. This is bad because it means that all the points have similar distance to each other, and so we cannot really figure out different labels for these points.\n", "\n", "Note: that the 2-norm has min value zero, and max value $\\sqrt{2}$, so for dimension 2, the maximum 2-norm is $\\sqrt{d}$, for dim=100, the maximum is $10$, etc, which is why the x-axis range changes from plot to plot.\n" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "scrolled": false }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "n = 1000\n", "dims = [2, 5, 10, 50, 100, 200, 1000, 2000, 10000]\n", "fig, axs = plt.subplots(3,3, figsize=(10,10))\n", "for i, ax in enumerate(axs.flat):\n", " d = dims[i]\n", " \n", " # randomly sample N points from [0,1]^dim\n", " X = np.random.uniform(low=0., high=1., size=(n, d))\n", " \n", " # compute pairwise distances between the points\n", " pdists = np.array([np.linalg.norm(X[i]-X[j], ord=2) for i in range(n) for j in range (i+1, n)])\n", " \n", " # plot histogram\n", " ax.hist(pdists, range=[0, np.sqrt(d)], bins=50) \n", " ax.set_title(f\"dim = {d}\")\n", "plt.suptitle(f\"Distribution of Parwise Distances for n={n} sampled points on unit cube\")\n", "plt.tight_layout()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Decision Trees for Classification\n", "An important nonparametric model is the decision tree, which is the building block for more complex models such as the random forest and adaptive boosting algorithms we will encounter later on in the course. Decision trees can be thought of as a large set of `if-then-else` statements. They are also referred to as recursive partitioning algorithms, since they repeatedly partition the input space into smaller and smaller subsets or regions. In this section we will focus on decision trees for classification, but note that they are also able to be used for regression problems." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Recursive Partitioning and visualizing classifiers\n", "We first aim to understand decision trees as a partitioning estimator. To do so, we'll need the following helper function that allows us to visualize the classifier, and you can use it as a black box for the remainder of the lab." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "# Helper function for visualising classifiers and decision surfaces \n", "def visualize_classifier(model, X, y, ax=None, cmap='rainbow', title=None):\n", " \n", " # reference: Python Data Science Handbook by Jake VanderPlas\n", " ax = ax or plt.gca()\n", " \n", " # Plot the training points\n", " \n", " if np.any(y==-1):\n", " y[y==-1] = 0. # fix to get scatter c=y arg working when we use -1,1 coding\n", " \n", " ax.scatter(X[:, 0], X[:, 1], c=y, s=30, cmap=cmap,\n", " clim=(y.min(), y.max()), zorder=3)\n", " ax.axis('tight')\n", " xlim = ax.get_xlim()\n", " ylim = ax.get_ylim()\n", " \n", " # compute predictions on grid\n", " xx, yy = np.meshgrid(np.linspace(*xlim, num=200),\n", " np.linspace(*ylim, num=200))\n", " Z = model.predict(np.c_[xx.ravel(), yy.ravel()]).reshape(xx.shape)\n", " \n", " if np.any(Z==-1): # fix to get c=y working\n", " Z[Z==-1] = 0\n", "\n", " # Create a color plot with the results\n", " n_classes = len(np.unique(y))\n", " contours = ax.contourf(xx, yy, Z, alpha=0.3,\n", " levels=np.arange(n_classes + 1) - 0.5,\n", " cmap=cmap, zorder=1)\n", "\n", " ax.set(xlim=xlim, ylim=ylim)\n", " if title:\n", " ax.set_title(title)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The `sklearn.datasets.make_blobs` function gives us a quick way to create toy data for classification. In the following we'll create a 3 class classification problem " ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "scrolled": true }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "from sklearn.datasets import make_blobs\n", "\n", "X, y = make_blobs(n_samples=120, # total number of samples \n", " centers=[[0,0], [0,2], [-2,1]], # cluster centers of the 3 classes\n", " random_state=123, # reproducibility \n", " cluster_std=0.6) # how spread out are the samples from their center\n", " \n", "plt.scatter(X[:, 0], X[:, 1], c=y, s=50) # scatter with color=label\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We will also use the `sklearn.neighbors.KNeighborsClassifier` and `sklearn.tree.DecisionTreeClassifier` objects to demonstrate the `visualize_classifier` function. The shaded regions correspond to how the classifier would classify a point that falls in that region. Note that for the kNN classifier, the regions are quite jagged, whereas the regions for the DT are always constructed by splitting using straight lines." ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn.neighbors import KNeighborsClassifier\n", "\n", "# kNN with 5 neighbours and euclidean distance (p=2)\n", "model0 = KNeighborsClassifier(n_neighbors=5, p=2) \n", "\n", "# DT with max depth of 3 and entropy criterion (this is information gain from lectures/tutorials)\n", "model1 = DecisionTreeClassifier(max_depth=3, criterion='entropy', random_state=3) \n", "\n", "# fit both models\n", "model0.fit(X,y)\n", "model1.fit(X,y)\n", "\n", "\n", "# visualize classifiers\n", "fig, axes = plt.subplots(1,2, figsize=(10,6))\n", "\n", "# scatter data\n", "axes[0].scatter(X[:, 0], X[:, 1], c=y, s=50)\n", "axes[1].scatter(X[:, 0], X[:, 1], c=y, s=50)\n", "\n", "# classifier plot\n", "visualize_classifier(model0, X, y, ax=axes[0], title=\"kNN Classifier\")\n", "visualize_classifier(model1, X, y, ax=axes[1], title=\"DT Classifier\")\n", "\n", "plt.tight_layout()\n", "plt.show()\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Another good way to visualize a tree is to look directly at the rules it uses to perform the splits at each depth. Recall that for a classification decision tree, we pass an input down the tree and look at the majority class in the child node that it falls into to find its predicted class. In regression, we do something identical, except that instead of taking a majority, we can use an average (or weighted average) of the points in the child node (similar to kNN regression)\n", "\n", "Note that the color of the labels correspond to the majority class at that particular point. The plot also gives us the distribution of the three classes at each node (`value`), and tells us the number of samples that falls into a particular node. We also can see the information gain (`entropy`) (see tutorials for details) for each node." ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "from sklearn import tree\n", "\n", "fig, axes = plt.subplots(1, 1,figsize = (3,3), dpi=300)\n", "tree.plot_tree(model1, # fitted decision tree\n", " feature_names=['f1', 'f2'], # some names for the features\n", " class_names=['t1', 't2', 't3'], # some names for the class labels\n", " filled=True)\n", "plt.show()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", " \n", "#### Exercise: \n", "Generate data using the following code:\n", " \n", "```\n", "X, y = make_blobs(n_samples=500, \n", " centers=[[0,0], [0,2], [-2,1], [-2,2], [3,3], [1,-2]], \n", " random_state=123, \n", " cluster_std=0.6) \n", "```\n", "Then fit a decision tree (using information gain for splits) with `max_depth` set to $1,2,\\dots, 12$ and visualize the classifier (use a $3 \\times 4$ grid). What do you observe? Why do you think decision trees are described as performing 'recursive partitioning'?" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "\n", "#### Extended Data Exercise\n", "In this exercise, we'll work with the `titanic.csv` dataset which you should have available in the same working directory where this notebook is located. If you get stuck at any point, the first thing to do will be to google your question - when it comes to data analysis in Python it is almost impossible to remember all the different commands, and https://stackoverflow.com/ is your best friend.\n", " \n", "1. Using `pandas`, load in the dataset and call the dataframe `df`. Run `df.info()` to gain an understanding of the different features.\n", "\n", "2. Remove the following columns `'PassengerId','Name', 'Ticket','Cabin', 'Embarked'` from the data, as we won't be using them for this exercise.\n", " \n", "3. Use the `dropna()` method to remove any missing rows.\n", " \n", " \n", "4. We would like to treat all attributes as numeric, so convert the `Sex` feature to numeric, and code males as 0 and females as 1. Hint: a good approach is to use the `df.Sex.map()` method.\n", " \n", "5. We will be interested in predicting `Survived` as a function of the remaining features. Create `X,y` `numPy` arrays to use for this problem. Be sure to also save a copy of the names of the columns. Hint: use `iloc` to index pandas dataframes, and use `to_numpy()` to convert pandas dataframes to numpy arrays.\n", " \n", "6. Create train/test datasets using 70\\% of your data for training.\n", " \n", "7. Fit decision trees to your training dataset: do this for `max_depth`=1,...,20. For each depth, record the train and test classification error (1-accuracy). Create a plot of error against depth, and plot both the train and test errors. What do you notice?\n", " \n", "8. Re-fit the model on the entire dataset with a depth 3 tree. Plot the decision tree and interpret the results." ] } ], "metadata": { "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": "3.7.4" } }, "nbformat": 4, "nbformat_minor": 5 }