The ultimate guide to data science interviews. Developed by FAANG engineers, it covers Python, algorithms, ML concepts, and interview questions. Get interview-ready in just a few hours.

DSTar.tar.gz

code-job

r-job

This course will increase your skills to crack the data science or machine learning interview. You will cover all the most common data science and ML concepts coupled with relevant interview questions.
You will start by covering Python basics as well as the most widely used algorithms and data structures. From there, you will move on to more advanced topics like feature engineering, unsupervised learning, as well as neural networks and deep learning.
This course takes a non-traditional approach to interview prep, in that it focuses on data science fundamentals instead of open-ended questions.

In all, this course will get you ready for data science interviews. By the time you finish this course, you will have reviewed all the major concepts in data science and will have a good idea of what interview questions you can expect.

Data Science Interview Handbook

We worked on parametric approaches in the previous lesson. This lesson is dedicated to the nonparametric approach for regression.

Using the appropriate value of parameters, we can fit the data. But what if fitting the data is not possible with one model? 

See the example below.


We can think of fitting a simple regression line on this data. It will look similar to this.

However, this is not ideal. Maybe we can try high order polynomials. See the example of fit below.

It is better than the previous solution, but it still does not capture the best model. What if we can fit the data locally (small points) instead of trying globally?

This is a good and reasonable fit. The only limitation is that we need a good amount of data to fit these local models.

<h2> Nearest-neighbor regression </h2>

In this technique, we predict the value based on past nearest data. We find the most similar K values to the new value and take their observations to predict the value by a simple average or weighted average. Consider the below example.

Assume that all the blue points are training data and we have to predict a green point. We can use the nearest-neighbor regression to predict the value. Consider K=1. So, we predict the value of the nearest target value of its neighbor.

This is called 1-nearest neighbor regression.

Instead of 1-nearest neighbor, K neighbors are considered. This will give a more generalized prediction and safety from outliers. We can also provide weights on neighbors to get weighted nearest neighbors.

<b> Distance metrics </b> 

We choose the neighbor based on the minimum distance from the new point. We use different types of metrics to get this distance. A few examples are correlation-based, rank-based, cosine similarity, Euclidean, etc.

What are the advantages of K-nearest neighbors (K>1) compared to 1-nearest neighbor?

<p>What are the advantages of K-nearest neighbors (K&gt;1) compared to 1-nearest neighbor?</p>


Quiz: Value of K in KNN

<p>Quiz: Value of K in KNN</p>


If we keep increasing data, what would the model complexity of the K-nearest neighbor be? 

<p>If we keep increasing data, what would the model complexity of the K-nearest neighbor be?</p>


Quiz: Complexity of KNN

<p>Quiz: Complexity of KNN</p>


We can use the Euclidean, Manhattan, and Minkowski distance matrices. If we have categorical data, we can use the Hamming distance.

<p>We can use the Euclidean, Manhattan, and Minkowski distance matrices. If we have categorical data, we can use the Hamming distance.</p>


What distance matrices can be used in K-nearest neighbor regression?

<p>What distance matrices can be used in K-nearest neighbor regression?</p>


This is based on the idea of feature similarity. If the new instance has features similar to the previous instance, it tends to behave similarly. That’s why we do the prediction based on the closest example.

<p>This is based on the idea of feature similarity. If the new instance has features similar to the previous instance, it tends to behave similarly. That’s why we do the prediction based on the closest example.</p>


What is the concept behind the nearest neighbor regression?

<p>What is the concept behind the nearest neighbor regression?</p>


<h3> <div style="margin-bottom: 10 px;
  padding: 2px 10px; background-color:#f6f6f6;
  border-left: 24px solid #000000">
  <p> Interview question: </h3>


<h2> Kernel regression </h2>

While we can weigh neighbors in the K-nearest neighbor regression. In kernel regression, we weigh all the data points in the dataset.

$$ \hat{y} = \frac{\sum_{i=1}^{i=N}C{ei}y_i}{\sum_{i=1}^{i=N}C{ei}} = \frac{\sum_{i=1}^{i=N}Kernel (distance(x_ix_e)*y_i}{\sum_{i=1}^{i=N}Kernel (distance(x_ix_e)}

Learn about the nonparametric regression techniques: Nearest neighbor regression and kernel regression.

Are You Ready to Become a Data Scientist?

Python Basics

Python Libraries

More Data Science Tools

Data Structures and Algorithms - I

Data Structures and Algorithms - II

Statistics and Probability

Feature Engineering

Basics of Machine Learning

Regression

Classification

Unsupervised Learning

Advanced Topics in Machine Learning

Conclusion

Nonparametric Regression