How to crack Data Science in 30 days?

The scope of Data Science is widening everyday and so is the amount of information available in this domain.

As more and more aspirants pursue Data Science and there is an information overload on the internet, it is very easy to get overwhelmed and confused as to where to start?

Another striking question that crosses many Data Science aspirants is the time-frame to learn Data Science. Many aspirants are working professionals strapped on time and can devote only a couple of months or less to their studies.

So the real question is “is it even possible to learn Data Science in 30 days?”

The more practical answer is No.. It is not possible to cover all depths of Data Science in merely 30 days but a strict schedule might help you scale this mission.

You will first need to be sure that you absolutely have to follow and stick to the 30 days plan to be successful in this venture.

30 Days Study Plan

The best way to pursue your 30 Day Data Science Plan is to divide the month into 4 weeks and dedicate each week to developing one skill each.

Here’s how your 30 Day Study Plan should look like..

Week 1:

In the first week, assuming that you are a beginner and you do not know anything about mathematics, you do not know a lot about programming, or might just know a little bit of programming and a little bit of math, but not as much as required in Data Science. 

In the first week, what you should do as a newbie is to make your Python fundamentals really strong.

In the first week, after getting your grip on Python, make sure to get your hands dirty in the basic Python code. So this is what you should focus on in the first 3-4 days. 

The second thing to focus on in the next 2-3 days is to get your statistics concepts right. 

So clearing the basic statistical concepts is something that will help you in the later part of Data Science and make the subsequent 3 weeks much easier.

A thorough knowledge of hypothesis testing, probability and inferential statistics and experience in working with statistical models.

When you use statistics, you will aid the stakeholders in designing and evaluating experiments and taking measured decisions. Skills for a Data Scientist include the need to be acquainted with statistical methods and techniques like maximum likelihood, distributions, estimators, Logistic Regression, Clustering and Linear Regression etc.

Once you have mastered the first two steps which is learning Python and learning statistics basics in the first week, other things will fall into place by themselves.

Here, we are to assume that you have blocked 30 days solely for the pursuit of Data Science.

Week 2: 

Week 2 would be all about getting your hands dirty in the Data Analysis part. In Data Analysis you need to be very comfortable to master Pandas.

Pandas is used for Data Transformation, it should take you a total of 1-2 days, you should spend the next couple of days with NumPy and subsequently 1-2 days you should spend getting familiar with data visualization packages like Matplotlib and Seaborn.

After this, you should bring all of these packages together as real-world Data Science projects, and implement all of these packages in a data set that you could possibly take from any source like UC Irvine datasets or Kaggle datasets and perform just a basic level of data analysis so that you are comfortable with everything that you have learned. 

Week 1 was about mastering the fundamentals of Python and statistics. Week 2 is all about Data Analysis that I’ve just discussed.

Pin

Week 3:

In Week 3, you should move to Machine Learning. 

Do not bother about learning 20-30 different Machine Learning algorithms. The best way to get out the most from a week would be to master 5-6 Machine Learning algorithms that are the building blocks of Machine Learning that you need to know for Data Science.

For example, I would spend one day understanding the very basics of Machine Learning, how is it applied, etc, I would spend Day 2 in Linear Regression, I would understand the theory behind Linear Regression, the math behind Linear Regression, the assumptions of Linear Regression, and I would also implement Linear Regression in a project. And I would also see how Linear Regression is used from Scikit Learn library of Python.

Here is a list of Machine Learning algorithms you can pick up from:

1. Linear regression 
2. Logistic Regression
3. Classification And Regression Tree 
4. Random Forest
5. Naive Bayes Classifier
6. KNN
7. Support Vector Machines
8. K-means
9. Linear Discriminant Analysis
10. Principal Component Analysis

This is a list of a handful of supervised Machine Learning algorithms and unsupervised Machine Learning algorithms that you should know and try to master in the third week of your 30 Day Schedule.

Week 4: 

In the fourth week, you should continue your journey into Machine Learning by learning some more unsupervised algorithms by devoting 2-3 days. 

The next 3 days to master Data Science, you should implement all your learnings to more data sets from different industries.

For example, pick up a credit card fraud detection dataset, pick up a telecom data set, pick up a simple data set that could help you implement some of these Machine Learning algorithms in the real-world scenario, right.

However, cracking machine learning is not limited to enrolling for classes, taking notes, appearing for tests and getting a certification. It is crucial to put all you’ve learned into practice, and we mean real-world practice!

So this is the four-week plan that you should follow if you are on a crash course, so to speak, to master Data Science.

Realistically, cracking Data Science in 30 days can only be about mastering the fundamentals. You need to practice more and more and over and over to sharpen your skills.

This plan is only for individuals who do not have the resources or time to invest in a more branched out and comprehensive plan for Data Science Learning. 

However, it is not advisable to limit yourself to a 30 day Data Science Study Plan because it is very difficult to stick to such a tight schedule. Data Science is a vast subject with many tools and skills to learn from and many developments happening in real-time.

A 30 day period to limit yourself is not so feasible as a Data Scientist. However, if due to any considerations you absolutely need a 30 day learning time-frame, it is best that you follow this time schedule.

Let us know in the comments section how you will tackle learning Data Science in a limited time period and how your study schedules look like.