Skip to content
Snippets Groups Projects
Commit b07ec1f9 authored by Zeynep Hakguder's avatar Zeynep Hakguder
Browse files

added learning curve

parent e9cccf6f
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
ProgrammingAssignment1-Solution.ipynb
+ 0
− 33
View file @ b07ec1f9
%% Cell type:markdown id: tags:
# k-Nearest Neighbor
%% Cell type:markdown id: tags:
You can use numpy for array operations and matplpotlib for plotting for this assignment. Please do not add other libraries.
%% Cell type:code id: tags:
``` python
import numpy as np
import matplotlib.pyplot as plt
```
%% Cell type:markdown id: tags:
Following code makes the Model class and relevant functions available from model.ipynb.
%% Cell type:code id: tags:
``` python
%run 'model-Solution.ipynb'
```
%% Cell type:markdown id: tags:
Choice of distance metric plays an important role in the performance of kNN. Let's start by implementing a distance method in the "distance" function below. It should take two data points and the name of the metric and return a scalar value.
%% Cell type:code id: tags:
``` python
def distance(x, y, metric):
'''
x: a 1xd array
y: a 1xd array
metric: Euclidean, Hamming, etc.
'''
#raise NotImplementedError
if metric == 'Euclidean':
dist = np.sum(np.square((x-y)))
####################################
return dist # scalar distance btw x and y
```
%% Cell type:markdown id: tags:
We can implement our kNN classifier. kNN class inherits Model class. Implement "fit" and "predict" methods. Use the "distance" function you defined above. "fit" method takes $k$ as an argument. "predict" takes as input the feature vector for a single test point and outputs the predicted class, and the proportion of predicted class labels in $k$ nearest neighbors.
%% Cell type:code id: tags:
``` python
class kNN(Model):
def fit(self, k, distance_f, **kwargs):
#raise NotImplementedError
self.k = k
self.distance_f = distance_f
self.distance_metric = kwargs['metric']
#######################
return
# vary the threshold value for ROC analysis
def predict(self, test_points):
chosen_labels = []
for test_point in self.features[test_indices]:
#raise NotImplementedError
tmp_dist = [np.inf] * self.k
distances = []
labels = []
for index in self.training_indices:
dist = self.distance_f(self.features[index], test_point, self.distance_metric)
distances.append(dist)
labels.append(self.labels[index])
a_order = np.argsort(distances)
tmp_labels = list(np.array(labels)[a_order[::-1]][:self.k])
b = tmp_labels.count(1)
chosen_labels.append(b/self.k)
##########################
# return the predicted class label and the following ratio:
# number of points that have the same label as the test point / k
return np.array(chosen_labels)
```
%% Cell type:markdown id: tags:
It's time to build and evaluate our model now. Remember you need to provide values to $p$, $v$ parameters for "partition" function and to $file\_path$ for "preprocess" function.
%% Cell type:code id: tags:
``` python
# populate the keyword arguments dictionary kwargs
kwargs = {'p': 0.3, 'v': 0.1, 'seed': 123, 'file_path': 'madelon_train'}
# initialize the model
my_model = kNN(preprocessor_f=preprocess, partition_f=partition, **kwargs)
```
%% Cell type:markdown id: tags:
Assign a value to $k$ and fit the kNN model. You do not need to change the value of the $threshold$ parameter yet.
%% Cell type:code id: tags:
``` python
kwargs_f = {'metric': 'Euclidean'}
my_model.fit(k = 10, distance_f=distance, **kwargs_f)
```
%% Cell type:markdown id: tags:
Evaluate your model on the test data and report your accuracy. Also, calculate and report the confidence interval on the generalization error estimate.
%% Cell type:code id: tags:
``` python
final_labels = my_model.predict(my_model.test_indices)
```
%% Cell type:markdown id: tags:
Now that we have the true labels and the predicted ones from our model, we can build a confusion matrix and see how accurate our model is. Implement the "conf_matrix" function that takes as input an array of true labels ($true$) and an array of predicted labels ($pred$). It should output a numpy.ndarray.
%% Cell type:code id: tags:
``` python
def conf_matrix(true_l, pred, threshold):
tp = tn = fp = fn = 0
for i in range(len(true_l)):
tmp = -1
if pred[i] > threshold:
tmp = 1
if tmp == true_l[i]:
if true_l[i] == 1:
tp += 1
else:
tn += 1
else:
if true_l[i] == 1:
fn += 1
else:
fp += 1
return np.array([tp,tn, fp, fn])
# returns the confusion matrix as numpy.ndarray
#raise NotImplementedError
```
%% Cell type:code id: tags:
``` python
# You should see array([ 196, 106, 193, 105]) with seed 123
conf_matrix(my_model.labels[my_model.test_indices], final_labels, threshold= 0.5)
```
%% Output
array([196, 106, 193, 105])
%% Cell type:markdown id: tags:
ROC curves are a good way to visualize sensitivity vs. 1-specificity for varying cut off points. Now, implement a "ROC" function that predicts the labels of the test set examples using different $threshold$ values in "fit" and plot the ROC curve. "ROC" takes a list containing different $threshold$ parameter values to try and returns (sensitivity, 1-specificity) pair for each $parameter$ value.
%% Cell type:code id: tags:
``` python
def ROC(true, pred, value_list):
'''
true: nx1 array of true labels for test set
pred: nx1 array of predicted labels for test set
Calculate sensitivity and 1-specificity for each point in value_list
Return two nX1 arrays: sens (for sensitivities) and spec_ (for 1-specificities)
'''
return sens, spec_
```
%% Cell type:markdown id: tags:
We can finally create the confusion matrix and plot the ROC curve for our kNN classifier.
%% Cell type:code id: tags:
``` python
# confusion matrix
conf_matrix(true_classes, predicted_classes)
```
%% Cell type:code id: tags:
``` python
# ROC curve
roc_sens, roc_spec_ = ROC(true_classes, predicted_classes, np.arange(0.1, 1.0, 0.1))
plt.plot(roc_sens, roc_spec_)
plt.show()
```
......
ProgrammingAssignment1.ipynb
+ 49
− 5
View file @ b07ec1f9
%% Cell type:markdown id: tags:
# $k$-Nearest Neighbor
We'll implement $k$-Nearest Neighbor ($k$-NN) algorithm for this assignment. We recommend using [Madelon](https://archive.ics.uci.edu/ml/datasets/Madelon) dataset, although it is not mandatory. If you choose to use a different dataset, it should meet the following criteria:
* dependent variable should be binary (suited for binary classification)
* number of features (attributes) should be at least 50
* number of examples (instances) should be at least 1,000
A skeleton of a general supervised learning model is provided in "model.ipynb". Please look through it and complete the "preprocess" and "partition" methods.
### Assignment Goals:
In this assignment, we will:
* learn to split a dataset into training/validation/test partitions
* use the validation dataset to find a good value for $k$
* Having found the "best" $k$, we'll obtain final performance measures:
* accuracy, generalization error and ROC curve
%% Cell type:markdown id: tags:
You can use numpy for array operations and matplotlib for plotting for this assignment. Please do not add other libraries.
%% Cell type:code id: tags:
``` python
import numpy as np
import matplotlib.pyplot as plt
```
%% Cell type:markdown id: tags:
Following code makes the Model class and relevant functions available from model.ipynb.
%% Cell type:code id: tags:
``` python
%run 'model.ipynb'
```
%% Cell type:markdown id: tags:
Choice of distance metric plays an important role in the performance of $k$-NN. Let's start with implementing a distance method in the "distance" function below. It should take two data points and the name of the metric and return a scalar value.
%% Cell type:code id: tags:
``` python
def distance(x, y, metric):
'''
x: a 1xd array
y: a 1xd array
metric: Euclidean, Hamming, etc.
'''
raise NotImplementedError
return dist # scalar distance btw x and y
```
%% Cell type:markdown id: tags:
### $k$-NN Class Methods
%% Cell type:markdown id: tags:
We can start implementing our $k$-NN classifier. $k$-NN class inherits Model class. You'll need to implement "fit" and "predict" methods. Use the "distance" function you defined above. "fit" method takes $k$ as an argument. "predict" takes as input an $mxd$ array containing $d$-dimensional $m$ feature vectors for examples and outputs the predicted class and the proportion of predicted class labels in $k$ nearest neighbors.
%% Cell type:code id: tags:
``` python
class kNN(Model):
'''
Inherits Model class. Implements the k-NN algorithm for classification.
'''
def fit(self, k, distance_f, **kwargs):
'''
Fit the model. This is pretty straightforward for k-NN.
'''
# set self.k, self.distance_f, self.distance_metric
raise NotImplementedError
return
def predict(self, test_indices):
raise NotImplementedError
pred = []
# for each point in test points
# use your implementation of distance function
# distance_f(..., distance_metric)
# to find the labels of k-nearest neighbors.
# Find the ratio of the positive labels
# and append to pred with pred.append(ratio).
return np.array(pred)
```
%% Cell type:markdown id: tags:
### Build and Evaluate the Model (Accuracy, Confidence Interval, Confusion Matrix)
%% Cell type:markdown id: tags:
It's time to build and evaluate our model now. Remember you need to provide values to $p$, $v$ parameters for "partition" function and to $file\_path$ for "preprocess" function.
%% Cell type:code id: tags:
``` python
# populate the keyword arguments dictionary kwargs
kwargs = {'p': 0.3, 'v': 0.1, seed: 123, 'file_path': 'madelon_train'}
# initialize the model
my_model = kNN(preprocessor_f=preprocess, partition_f=partition, **kwargs)
```
%% Cell type:markdown id: tags:
Assign a value to $k$ and fit the kNN model.
%% Cell type:code id: tags:
``` python
kwargs_f = {'metric': 'Euclidean'}
my_model.fit(k = 10, distance_f=distance, **kwargs_f)
```
%% Cell type:markdown id: tags:
Evaluate your model on the test data and report your accuracy. Also, calculate and report the confidence interval on the generalization error estimate.
Evaluate your model on the test data and report your **accuracy**. Also, calculate and report the confidence interval on the generalization **error** estimate.
%% Cell type:code id: tags:
``` python
final_labels = my_model.predict(my_model.test_indices)
# Calculate accuracy and generalization error with confidence interval here.
# For now, We will consider a data point as predicted in the positive class if more than 0.5
# of its k-neighbors are positive.
threshold = 0.5
```
%% Output
---------------------------------------------------------------------------
NameError Traceback (most recent call last)
<ipython-input-3-e365162558f6> in <module>()
----> 1 final_labels = my_model.predict(my_model.test_indices)
2 threshold = 0.5
3 # Calculate accuracy and generalization error with confidence interval here. For now, We will consider a data point as predicted in the positive class if more than 0.5 of its k-neighbors are positive.
NameError: name 'my_model' is not defined
%% Cell type:markdown id: tags:
# TODO: learning curve
### Plotting a learning curve
A learning curve shows how error changes as the training set size increases. For more information, see [learning curves](https://www.dataquest.io/blog/learning-curves-machine-learning/).
We'll plot the error values for training and validation data while varying the size of the training set.
%% Cell type:code id: tags:
``` python
training_sizes = np.xrange(0, my_model.train_size + 1, 100)
# Calculate error for each entry in training_sizes
# for training and validation sets and populate
# error_train and error_val arrays. Each entry in these arrays
# should correspond to each entry in training_sizes.
plt.plot(training_sizes, error_train, 'r', label = 'training_error')
plt.plot(training_sizes, error_val, 'g', label = 'validation_error')
plt.legend()
plt.show()
```
%% Cell type:markdown id: tags:
### Computing the confusion matrix for $k = 10$
Now that we have the true labels and the predicted ones from our model, we can build a confusion matrix and see how accurate our model is. Implement the "conf_matrix" function (in model.ipynb) that takes as input an array of true labels ($true$) and an array of predicted labels ($pred$). It should output a numpy.ndarray. You do not need to change the value of the threshold parameter yet.
%% Cell type:code id: tags:
``` python
# You should see array([ 196, 106, 193, 105]) with seed 123
conf_matrix(my_model.labels[my_model.test_indices], final_labels, threshold= 0.5)
```
%% Cell type:markdown id: tags:
### Finding a good value for $k$
We can use the validation set to come up with a $k$ value that results in better performance in terms of accuracy. Additionally, in some cases, predicting examples from a certain class correctly is more critical than other classes. In those cases, we can use the confusion matrix to find a good trade off between correct and wrong predictions and allow more wrong predictions in some classes to predict more examples correctly in a that class.
Below calculate the accuracies and confusion matrices for different values of $k$ using the validation set. Report a good $k$ value and use it in the analyses that follow this section.
%% Cell type:code id: tags:
``` python
# Change values of $k.
# Calculate accuracies for the validation set.
# Report a good k value that you'll use in the following analyses.
```
%% Cell type:markdown id: tags:
### ROC curve and confusion matrix for the final model
ROC curves are a good way to visualize sensitivity vs. 1-specificity for varying cut off points. Now, implement a "ROC" function that predicts the labels of the test set examples using different $threshold$ values in "predict" and plot the ROC curve. "ROC" takes a list containing different $threshold$ parameter values to try and returns two arrays; one where each entry is the sensitivity at a given threshold and the other where entries are 1-specificities.
%% Cell type:code id: tags:
``` python
def ROC(model, indices, value_list):
'''
model: a fitted k-NN model
indices: for data points to predict
value_list: array containing different threshold values
Calculate sensitivity and 1-specificity for each point in value_list
Return two nX1 arrays: sens (for sensitivities) and spec_ (for 1-specificities)
'''
# use predict_batch to obtain predicted labels at different threshold values
raise NotImplementedError
return sens, spec_
```
%% Cell type:markdown id: tags:
We can finally create the confusion matrix and plot the ROC curve for our optimal $k$-NN classifier.
%% Cell type:code id: tags:
``` python
# confusion matrix
conf_matrix(true_classes, predicted_classes)
```
%% Cell type:code id: tags:
``` python
# ROC curve
roc_sens, roc_spec_ = ROC(my_model, my_model.test_indices, np.arange(0.1, 1.0, 0.1))
plt.plot(roc_sens, roc_spec_)
plt.show()
```
......
model.ipynb
+ 8
− 3
View file @ b07ec1f9
%% Cell type:markdown id: tags:
# JUPYTER NOTEBOOK TIPS
Each rectangular box is called a cell.
* ctrl+ENTER evaluates the current cell; if it contains Python code, it runs the code, if it contains Markdown, it returns rendered text.
* alt+ENTER evaluates the current cell and adds a new cell below it.
* If you click to the left of a cell, you'll notice the frame changes color to blue. You can erase a cell by hitting 'dd' (that's two "d"s in a row) when the frame is blue.
%% Cell type:markdown id: tags:
# Supervised Learning Model Skeleton
We'll use this skeleton for implementing different supervised learning algorithms. Please complete "preprocess" and "partition" methods below.
%% Cell type:markdown id: tags:
You might need to preprocess your dataset depending on which dataset you are using. This step is for reading the dataset and for extracting features and labels. The "preprocess" function should return an $n \times d$ features array, and an $n \times 1$ labels array, where $n$ is the number of examples and $d$ is the number of features in the dataset. In cases where there is a big difference between the scales of feautures, we want to normalize the features to have values in the same range [0,1]. If that is the case with your dataset, output normalized features to get better prediction results.
%% Cell type:code id: tags:
``` python
def preprocess(file_path):
'''
file_path: where to read the dataset from
returns nxd features, nx1 labels
'''
# You might find np.genfromtxt useful for reading in the file. Be careful with the file delimiter,
# e.g. for comma-separated files use delimiter=',' argument.
raise NotImplementedError
return features, labels
```
%% Cell type:markdown id: tags:
Next, you'll need to split your dataset into training and validation and test sets. The "split" function should take as input the size of the whole dataset and randomly sample a proportion $p$ of the dataset as test partition and a proportion of $v$ as validation partition. The remaining will be used as training data. For example, to keep 30% of the examples as test and %10 as validation, set $p=0.3$ and $v=0.1$. You should choose these values according to the size of the data available to you. The "split" function should return indices of the training, validation and test sets. These will be used to index into the whole training set.
%% Cell type:code id: tags:
``` python
def partition(size, p, v):
'''
size: number of examples in the whole dataset
p: proportion kept for test
v: proportion kept for validation
'''
# np.random.choice might come in handy. Do not sample with replacement!
# Be sure to not use the same indices in test and validation sets!
# use the first np.ceil(size*p) for test,
# the following np.ceil(size*v) for validation set.
raise NotImplementedError
# return two 1d arrays: one keeping validation set indices, the other keeping test set indices
return val_indices, test_indices
```
%% Cell type:code id: tags:
``` python
class Model:
# set the preprocessing function, partition_function
# use kwargs to pass arguments to preprocessor_f and partition_f
# kwargs is a dictionary and should contain p, v and file_path
# e.g. {'p': 0.3, 'v': 0.1, 'file_path': some_path}
def __init__(self, preprocessor_f, partition_f, **kwargs):
self.features, self.labels = preprocessor_f(kwargs['file_path'])
self.size = len(self.labels) # number of examples in dataset
self.feat_dim = self.features.shape[1] # number of features
self.val_indices, self.test_indices = partition_f(self.size, kwargs['p'], kwargs['v'])
self.val_size = len(self.val_indices)
self.test_size = len(self.test_indices)
self.train_indices = np.delete(np.arange(self.size), np.append(self.test_indices, self.val_indices), 0)
self.train_size = len(self.train_indices)
def fit(self):
raise NotImplementedError
def predict(self, testpoint):
raise NotImplementedError
```
%% Cell type:markdown id: tags:
## General supervised learning related functions
%% Cell type:markdown id: tags:
Implement the "conf_matrix" function that takes as input an array of true labels ($true$) and an array of predicted labels ($pred$). It should output a numpy.ndarray.
%% Cell type:code id: tags:
``` python
def conf_matrix(true, pred):
'''
true: nx1 array of true labels for test set
pred: nx1 array of predicted labels for test set
'''
raise NotImplementedError
tp = tn = fp = fn = 0
# calculate true positives (tp), true negatives(tn)
# false positives (fp) and false negatives (fn)
# returns the confusion matrix as numpy.ndarray
return np.array([tp,tn, fp, fn])
```
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment