How to use nearest neighbours for Regression?
MACHINE LEARNING RECIPES DATA CLEANING PYTHON DATA MUNGING PANDAS CHEATSHEET     ALL TAGS

How to use nearest neighbours for Regression?

How to use nearest neighbours for Regression?

This recipe helps you use nearest neighbours for Regression

Recipe Objective

Many a times while working on a dataset and using a Machine Learning model we don"t know which set of hyperparameters will give us the best result. Passing all sets of hyperparameters manually through the model and checking the result might be a hectic work and may not be possible to do.

To get the best set of hyperparameters we can use Grid Search. Grid Search passes all combinations of hyperparameters one by one into the model and check the result. Finally it gives us the set of hyperparemeters which gives the best result after passing in the model.

So this recipe is a short example of how can can use nearest neighbours for Regression.

Step 1 - Import the library

from sklearn import decomposition, datasets from sklearn import neighbors from sklearn.pipeline import Pipeline from sklearn.model_selection import GridSearchCV, cross_val_score from sklearn.preprocessing import StandardScaler

Here we have imported various modules like decomposition, datasets, tree, Pipeline, StandardScaler and GridSearchCV from differnt libraries. We will understand the use of these later while using it in the in the code snipet.
For now just have a look on these imports.

Step 2 - Setup the Data

Here we have created a regression dataset with python datasets. dataset = datasets.make_regression(n_samples=1000, n_features=20, n_informative=10, n_targets=1, bias=0.0, effective_rank=None, tail_strength=0.5, noise=0.0, shuffle=True, coef=False, random_state=None) X = dataset[0] y = dataset[1]

Step 3 - Using StandardScaler and PCA

StandardScaler is used to remove the outliners and scale the data by making the mean of the data 0 and standard deviation as 1. So we are creating an object std_scl to use standardScaler. std_slc = StandardScaler()

We are also using Principal Component Analysis(PCA) which will reduce the dimension of features by creating new features which have most of the varience of the original data. pca = decomposition.PCA()

Here, we are using KNeighbors Regressor as a Machine Learning model to use GridSearchCV. So we have created an object KNN. KNN = neighbors.KNeighborsRegressor()

Step 5 - Using Pipeline for GridSearchCV

Pipeline will helps us by passing modules one by one through GridSearchCV for which we want to get the best parameters. So we are making an object pipe to create a pipeline for all the three objects std_scl, pca and knn. pipe = Pipeline(steps=[("std_slc", std_slc), ("pca", pca), ("KNN", KNN)])

Now we have to define the parameters that we want to optimise for these three objects.
StandardScaler doesnot requires any parameters to be optimised by GridSearchCV.
Principal Component Analysis requires a parameter "n_components" to be optimised. "n_components" signifies the number of components to keep after reducing the dimension. n_components = list(range(1,X.shape[1]+1,1))

DecisionTreeClassifier requires two parameters "n_neighbors" and "algorithm" to be optimised by GridSearchCV. So we have set these two parameters as a list of values form which GridSearchCV will select the best value of parameter. n_neighbors = [2, 3, 5, 10] algorithm = ["auto", "ball_tree", "kd_tree", "brute"]

Now we are creating a dictionary to set all the parameters options for different objects. parameters = dict(pca__n_components=n_components, KNN__n_neighbors=n_neighbors, KNN__algorithm=algorithm)

Step 6 - Using GridSearchCV and Printing Results

Before using GridSearchCV, lets have a look on the important parameters.

  • estimator: In this we have to pass the models or functions on which we want to use GridSearchCV
  • param_grid: Dictionary or list of parameters of models or function in which GridSearchCV have to select the best.
  • Scoring: It is used as a evaluating metric for the model performance to decide the best hyperparameters, if not especified then it uses estimator score.
Making an object clf for GridSearchCV and fitting the dataset i.e X and y clf = GridSearchCV(pipe, parameters) clf.fit(X, y) Now we are using print statements to print the results. It will give the values of hyperparameters as a result. print("Best Number Of Components:", clf.best_estimator_.get_params()["pca__n_components"]) print(); print(clf.best_estimator_.get_params()["KNN"]) CV_Result = cross_val_score(clf, X, y, cv=3, n_jobs=-1, scoring="r2") print(); print(CV_Result) print(); print(CV_Result.mean()) print(); print(CV_Result.std()) As an output we get:

Best Number Of Components: 20

KNeighborsRegressor(algorithm="auto", leaf_size=30, metric="minkowski",
          metric_params=None, n_jobs=None, n_neighbors=10, p=2,
          weights="uniform")

[0.60800965 0.53874633 0.57159348]

0.5727831547316445

0.028289142973677704

Download Materials

Relevant Projects

Resume parsing with Machine learning - NLP with Python OCR and Spacy
In this machine learning resume parser example we use the popular Spacy NLP python library for OCR and text classification.

RASA NLU chatbot creation
The project will use rasa NLU for the Intent classifier, spacy for entity tagging, and mongo dB as the DB. The project will incorporate slot filling and context management and will be supporting the following intent and entities. Intents : product_info | ask_price|cancel_order Entities : product_name|location|order id The project will demonstrate how to generate data on the fly, annotate using framework and how to process those for different pieces of training as discussed above .

Abstractive Text Summarization using Transformers-BART Model
Deep Learning Project to implement an Abstractive Text Summarizer using Google's Transformers-BART Model to generate news article headlines.

Ensemble Machine Learning Project - All State Insurance Claims Severity Prediction
In this ensemble machine learning project, we will predict what kind of claims an insurance company will get. This is implemented in python using ensemble machine learning algorithms.

Grouping similar schools/colleges using scorecard and other factors
Use cluster analysis to identify the groups of characteristically similar schools in the College Scorecard dataset. Considerations: Clustering Algorithm Data Preparation How will you deal with missing values? Categorical variables? Feature intercorrelations? Feature normalization or scaling? Dimensionality reduction? Hyperparameters How will you set the parameters -- the algorithm's knobs and dials, so to speak -- in order to achieve valid and useful output? Interpretation Is it possible to explain what each cluster represents? Did you retain or prepare a set of features that enables a meaningful interpretation of the clusters? Do the compositions of the clusters seem to make sense? Validation How will you measure the validity of your clustering process? Which metrics will you use and how will you apply them?

Time Series Analysis Project in R on Stock Market forecasting
In this time series project, you will build a model to predict the stock prices and identify the best time series forecasting model that gives reliable and authentic results for decision making.

Churn Prediction in Telecom using Machine Learning in R
Estimating churners before they discontinue using a product or service is extremely important. In this ML project, you will develop a churn prediction model in telecom to predict customers who are most likely subject to churn.

NLP and Deep Learning For Fake News Classification in Python
In this project you will use Python to implement various machine learning methods( RNN, LSTM, GRU) for fake news classification.

Data Science Project - Instacart Market Basket Analysis
Data Science Project - Build a recommendation engine which will predict the products to be purchased by an Instacart consumer again.

Digit Recognition using CNN for MNIST Dataset in Python
In this deep learning project, you will build a convolutional neural network using MNIST dataset for handwritten digit recognition.