{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"# HIDDEN\n",
"import numpy as np\n",
"%matplotlib inline\n",
"import matplotlib.pyplot as plt\n",
"plt.style.use('fivethirtyeight')\n",
"import pandas as pd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Implementing the Classifier ###\n",
"\n",
"We are now ready to implement a $k$-nearest neighbor classifier based on multiple attributes. We have used only two attributes so far, for ease of visualization. But usually predictions will be based on many attributes. Here is an example that shows how multiple attributes can be better than pairs."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Banknote authentication\n",
"\n",
"This time we'll look at predicting whether a banknote (e.g., a \\$20 bill) is counterfeit or legitimate. Researchers have put together a data set for us, based on photographs of many individual banknotes: some counterfeit, some legitimate. They computed a few numbers from each image, using techniques that we won't worry about for this course. So, for each banknote, we know a few numbers that were computed from a photograph of it as well as its class (whether it is counterfeit or not). Let's load it into a table and take a look.\n",
"\n",
"If you are running on your laptop, you should download the\n",
"[banknotes]({{ site.baseurl }}/data/banknotes.csv) file to the\n",
"same directory as this notebook."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"banknotes.plot.scatter('WaveletVar',\n",
" 'WaveletCurt',\n",
" c=banknotes['Color']);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Pretty interesting! Those two measurements do seem helpful for predicting whether the banknote is counterfeit or not. However, in this example you can now see that there is some overlap between the blue cluster and the gold cluster. This indicates that there will be some images where it's hard to tell whether the banknote is legitimate based on just these two numbers. Still, you could use a $k$-nearest neighbor classifier to predict the legitimacy of a banknote.\n",
"\n",
"Take a minute and think it through: Suppose we used $k=11$ (say). What parts of the plot would the classifier get right, and what parts would it make errors on? What would the decision boundary look like?\n",
"\n",
"The patterns that show up in the data can get pretty wild. For instance, here's what we'd get if used a different pair of measurements from the images:"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"banknotes.plot.scatter('WaveletSkew', 'Entropy',\n",
" c=banknotes['Color']);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"There does seem to be a pattern, but it's a pretty complex one. Nonetheless, the $k$-nearest neighbors classifier can still be used and will effectively \"discover\" patterns out of this. This illustrates how powerful machine learning can be: it can effectively take advantage of even patterns that we would not have anticipated, or that we would have thought to \"program into\" the computer."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Multiple attributes\n",
"\n",
"So far I've been assuming that we have exactly 2 attributes that we can use to help us make our prediction. What if we have more than 2? For instance, what if we have 3 attributes?\n",
"\n",
"Here's the cool part: you can use the same ideas for this case, too. All you have to do is make a 3-dimensional scatterplot, instead of a 2-dimensional plot. You can still use the $k$-nearest neighbors classifier, but now computing distances in 3 dimensions instead of just 2. It just works. Very cool!\n",
"\n",
"In fact, there's nothing special about 2 or 3. If you have 4 attributes, you can use the $k$-nearest neighbors classifier in 4 dimensions. 5 attributes? Work in 5-dimensional space. And no need to stop there! This all works for arbitrarily many attributes; you just work in a very high dimensional space. It gets wicked-impossible to visualize, but that's OK. The computer algorithm generalizes very nicely: all you need is the ability to compute the distance, and that's not hard. Mind-blowing stuff!\n",
"\n",
"For instance, let's see what happens if we try to predict whether a banknote is counterfeit or not using 3 of the measurements, instead of just 2. Here's what you get:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"from mpl_toolkits.mplot3d import Axes3D\n",
"ax = plt.figure(figsize=(8,8)).add_subplot(111, projection='3d')\n",
"ax.scatter(banknotes['WaveletSkew'],\n",
" banknotes['WaveletVar'],\n",
" banknotes['WaveletCurt'],\n",
" c=banknotes['Color']);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Awesome! With just 2 attributes, there was some overlap between the two clusters (which means that the classifier was bound to make some mistakes for pointers in the overlap). But when we use these 3 attributes, the two clusters have almost no overlap. In other words, a classifier that uses these 3 attributes will be more accurate than one that only uses the 2 attributes.\n",
"\n",
"This is a general phenomenon in classification. Each attribute can potentially give you new information, so more attributes sometimes helps you build a better classifier. Of course, the cost is that now we have to gather more information to measure the value of each attribute, but this cost may be well worth it if it significantly improves the accuracy of our classifier.\n",
"\n",
"To sum up: you now know how to use $k$-nearest neighbor classification to predict the answer to a yes/no question, based on the values of some attributes, assuming you have a training set with examples where the correct prediction is known. The general roadmap is this:\n",
"\n",
"1. identify some attributes that you think might help you predict the answer to the question.\n",
"2. Gather a training set of examples where you know the values of the attributes as well as the correct prediction.\n",
"3. To make predictions in the future, measure the value of the attributes and then use $k$-nearest neighbor classification to predict the answer to the question."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Distance in Multiple Dimensions ###\n",
"\n",
"We know how to compute distance in 2-dimensional space. If we have a point at coordinates $(x_0,y_0)$ and another at $(x_1,y_1)$, the distance between them is\n",
"\n",
"$$D = \\sqrt{(x_0-x_1)^2 + (y_0-y_1)^2}.$$\n",
"\n",
"In 3-dimensional space, the points are $(x_0, y_0, z_0)$ and $(x_1, y_1, z_1)$, and the formula for the distance between them is\n",
"\n",
"$$\n",
"D = \\sqrt{(x_0-x_1)^2 + (y_0-y_1)^2 + (z_0-z_1)^2}\n",
"$$\n",
"\n",
"In $n$-dimensional space, things are a bit harder to visualize, but I think you can see how the formula generalized: we sum up the squares of the differences between each individual coordinate, and then take the square root of that."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In the last section, we defined the function `distance` which returned the distance between two points. We used it in two-dimensions, but the great news is that the function doesn't care how many dimensions there are! It just subtracts the two arrays of coordinates (no matter how long the arrays are), squares the differences and adds up, and then takes the square root. To work in multiple dimensions, we don't have to change the code at all."
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"def distance(point1, point2):\n",
" \"\"\"Returns the distance between point1 and point2\n",
" where each argument is an array \n",
" consisting of the coordinates of the point\"\"\"\n",
" return np.sqrt(np.sum((point1 - point2)**2))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's use this on a [new dataset](https://archive.ics.uci.edu/ml/datasets/Wine). The table `wine` contains the chemical composition of 178 different Italian wines. The classes are the grape species, called cultivars. There are three classes but let's just see whether we can tell Class 1 apart from the other two.\n",
"\n",
"You can download the file from\n",
"[wine.csv]({{ site.baseurl }}/data/wine.csv)."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"wine_with_colors.plot.scatter('Flavanoids', 'Alcohol',\n",
" c=wine_with_colors['Color']);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The blue points (Class 1) are almost entirely separate from the gold ones. That is one indication of why the distance between two Class 1 wines would be smaller than the distance between wines of two different classes. We can see a similar phenomenon with a different pair of attributes too:"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"wine_with_colors.plot.scatter('Alcalinity of Ash', 'Ash',\n",
" c=wine_with_colors['Color']);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"But for some pairs the picture is more murky."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"wine_with_colors.plot.scatter('Magnesium', 'Total Phenols',\n",
" c=wine_with_colors['Color']);"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's see if we can implement a classifier based on all of the attributes. After that, we'll see how accurate it is."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### A Plan for the Implementation ###\n",
"\n",
"It's time to write some code to implement the classifier. The input is a `point` that we want to classify. The classifier works by finding the $k$ nearest neighbors of `point` from the training set. So, our approach will go like this:\n",
"\n",
"1. Find the closest $k$ neighbors of `point`, i.e., the $k$ wines from the training set that are most similar to `point`.\n",
"\n",
"2. Look at the classes of those $k$ neighbors, and take the majority vote to find the most-common class of wine. Use that as our predicted class for `point`.\n",
"\n",
"So that will guide the structure of our Python code."
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [],
"source": [
"def closest(training, p, k):\n",
" \"\"\" Find the closest k neighbors of p \"\"\"\n",
"\n",
"def majority(topkclasses):\n",
" \"\"\" Return majority vote from top k classes \"\"\"\n",
"\n",
"def classify(training, p, k):\n",
" kclosest = closest(training, p, k)\n",
" kclosest.classes = kclosest.select('Class')\n",
" return majority(kclosest)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Implementation Step 1 ###\n",
"\n",
"To implement the first step for the kidney disease data, we had to compute the distance from each patient in the training set to `point`, sort them by distance, and take the $k$ closest patients in the training set.\n",
"\n",
"That's what we did in the previous section with the point corresponding to Alice. Let's generalize that code. We'll redefine `distance` here, just for convenience."
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [],
"source": [
"def distance(point1, point2):\n",
" \"\"\"Returns the distance between point1 and point2\n",
" where each argument is an array\n",
" consisting of the coordinates of the point\"\"\"\n",
" return np.sqrt(np.sum((point1 - point2)**2))\n",
"\n",
"def all_distances(training, new_point):\n",
" \"\"\"Returns an array of distances\n",
" between each point in the training set\n",
" and the new point (which is a row of attributes)\"\"\"\n",
" attributes = training.drop(columns='Class')\n",
" def distance_from_point(row):\n",
" return distance(np.array(new_point), np.array(row))\n",
" return attributes.apply(distance_from_point, axis=1)\n",
"\n",
"def table_with_distances(training, new_point):\n",
" \"\"\"Augments the training table\n",
" with a column of distances from new_point\"\"\"\n",
" out = training.copy()\n",
" out['Distance'] = all_distances(training, new_point)\n",
" return out\n",
"\n",
"def closest(training, new_point, k):\n",
" \"\"\"Returns a table of the k rows of the augmented table\n",
" corresponding to the k smallest distances\"\"\"\n",
" with_dists = table_with_distances(training, new_point)\n",
" sorted_by_distance = with_dists.sort_values('Distance')\n",
" topk = sorted_by_distance.iloc[:k]\n",
" return topk"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Let's see how this works on our `wine` data. We'll just take the first wine and find its five nearest neighbors among all the wines. Remember that since this wine is part of the dataset, it is its own nearest neighbor. So we should expect to see it at the top of the list, followed by four others.\n",
"\n",
"First let's extract its attributes:"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [],
"source": [
"special_wine = wine.drop(columns='Class').iloc[0]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"And now let's find its 5 nearest neighbors."
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
Class
\n",
"
Alcohol
\n",
"
Malic Acid
\n",
"
Ash
\n",
"
Alcalinity of Ash
\n",
"
Magnesium
\n",
"
Total Phenols
\n",
"
Flavanoids
\n",
"
Nonflavanoid phenols
\n",
"
Proanthocyanins
\n",
"
Color Intensity
\n",
"
Hue
\n",
"
OD280/OD315 of diulted wines
\n",
"
Proline
\n",
"
Distance
\n",
"
\n",
" \n",
" \n",
"
\n",
"
0
\n",
"
1
\n",
"
14.23
\n",
"
1.71
\n",
"
2.43
\n",
"
15.6
\n",
"
127
\n",
"
2.80
\n",
"
3.06
\n",
"
0.28
\n",
"
2.29
\n",
"
5.64
\n",
"
1.04
\n",
"
3.92
\n",
"
1065
\n",
"
0.000000
\n",
"
\n",
"
\n",
"
54
\n",
"
1
\n",
"
13.74
\n",
"
1.67
\n",
"
2.25
\n",
"
16.4
\n",
"
118
\n",
"
2.60
\n",
"
2.90
\n",
"
0.21
\n",
"
1.62
\n",
"
5.85
\n",
"
0.92
\n",
"
3.20
\n",
"
1060
\n",
"
10.392805
\n",
"
\n",
"
\n",
"
45
\n",
"
1
\n",
"
14.21
\n",
"
4.04
\n",
"
2.44
\n",
"
18.9
\n",
"
111
\n",
"
2.85
\n",
"
2.65
\n",
"
0.30
\n",
"
1.25
\n",
"
5.24
\n",
"
0.87
\n",
"
3.33
\n",
"
1080
\n",
"
22.340748
\n",
"
\n",
"
\n",
"
48
\n",
"
1
\n",
"
14.10
\n",
"
2.02
\n",
"
2.40
\n",
"
18.8
\n",
"
103
\n",
"
2.75
\n",
"
2.92
\n",
"
0.32
\n",
"
2.38
\n",
"
6.20
\n",
"
1.07
\n",
"
2.75
\n",
"
1060
\n",
"
24.760232
\n",
"
\n",
"
\n",
"
46
\n",
"
1
\n",
"
14.38
\n",
"
3.59
\n",
"
2.28
\n",
"
16.0
\n",
"
102
\n",
"
3.25
\n",
"
3.17
\n",
"
0.27
\n",
"
2.19
\n",
"
4.90
\n",
"
1.04
\n",
"
3.44
\n",
"
1065
\n",
"
25.094663
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Class Alcohol Malic Acid Ash Alcalinity of Ash Magnesium \\\n",
"0 1 14.23 1.71 2.43 15.6 127 \n",
"54 1 13.74 1.67 2.25 16.4 118 \n",
"45 1 14.21 4.04 2.44 18.9 111 \n",
"48 1 14.10 2.02 2.40 18.8 103 \n",
"46 1 14.38 3.59 2.28 16.0 102 \n",
"\n",
" Total Phenols Flavanoids Nonflavanoid phenols Proanthocyanins \\\n",
"0 2.80 3.06 0.28 2.29 \n",
"54 2.60 2.90 0.21 1.62 \n",
"45 2.85 2.65 0.30 1.25 \n",
"48 2.75 2.92 0.32 2.38 \n",
"46 3.25 3.17 0.27 2.19 \n",
"\n",
" Color Intensity Hue OD280/OD315 of diulted wines Proline Distance \n",
"0 5.64 1.04 3.92 1065 0.000000 \n",
"54 5.85 0.92 3.20 1060 10.392805 \n",
"45 5.24 0.87 3.33 1080 22.340748 \n",
"48 6.20 1.07 2.75 1060 24.760232 \n",
"46 4.90 1.04 3.44 1065 25.094663 "
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"closest(wine, special_wine, 5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Bingo! The first row is the nearest neighbor, which is itself – there's a 0 in the `Distance` column as expected. All five nearest neighbors are of Class 1, which is consistent with our earlier observation that Class 1 wines appear to be clumped together in some dimensions."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Implementation Steps 2 and 3 ###\n",
"\n",
"Next we need to take a \"majority vote\" of the nearest neighbors and assign our point the same class as the majority."
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [],
"source": [
"def majority(topkclasses):\n",
" ones = np.count_nonzero(topkclasses == 1)\n",
" zeros = np.count_nonzero(topkclasses == 0)\n",
" if ones > zeros:\n",
" return 1\n",
" else:\n",
" return 0\n",
"\n",
"def classify(training, new_point, k):\n",
" closestk = closest(training, new_point, k)\n",
" return majority(closestk['Class'])"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"1"
]
},
"execution_count": 23,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"classify(wine, special_wine, 5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"If we change `special_wine` to be the last one in the dataset, is our classifier able to tell that it's in Class 0?"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0"
]
},
"execution_count": 24,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"special_wine = wine.drop(columns='Class').iloc[177]\n",
"classify(wine, special_wine, 5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Yes! The classifier gets this one right too.\n",
"\n",
"But we don't yet know how it does with all the other wines, and in any case we know that testing on wines that are already part of the training set might be over-optimistic. In the final section of this chapter, we will separate the wines into a training and test set and then measure the accuracy of our classifier on the test set."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"{% data8page Implementing_the_Classifier %}"
]
}
],
"metadata": {
"anaconda-cloud": {},
"jupytext": {
"metadata_filter": {
"notebook": {
"additional": "all",
"excluded": [
"language_info"
]
}
},
"text_representation": {
"extension": ".Rmd",
"format_name": "rmarkdown",
"format_version": "1.0",
"jupytext_version": "0.8.6"
}
},
"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": "3.7.2"
}
},
"nbformat": 4,
"nbformat_minor": 2
}