Business leaders are constantly thinking of how to innovate their products and offer better services than their competitor’s in the market. Data is one such term that has been trending for the past few years and a lot of students are wanting to pursue a career in the field of data science. The most obvious question that arises is How to become a data analyst ? or How to start a career in the field of data science? With the availability of high-speed internet today, real-time data figures are being created constantly. At the beginning of 2020, the digital universe was estimated to consist of 44 zettabytes of data. By 2025, approximately 463 exabytes would be created every 24 hours worldwide. Such statistics are being used by businesses to analyze and get useful insights to leverage business projects. This transformation of raw unstructured figures to sensible useful information can be done by data science. With the growth of the IT industry in India, the number of data analyst jobs, and data warehousing jobs are growing each day.

Steps to how to start a career in data science:

· STATISTICS:

The question arises why do we need statistics for handling data?

Math and statistics form a building block for machine learning algorithms. The statistic is used to derive meaningful insights from the transformed data. It helps in the process of proper collection of figures, applying correct analysis to the gathered figures, and effectively presenting the results.

So what do we need to study in statistics for a data scientist career path:

1. Types: Numerical / Categorical.

2. Measures of Central Tendency: Mean, Median, and Mode.

3. The measure of Variability: Range, Variance, Standard Deviations, Z-score, R -square, adjusted R square.

4. The measure of the relationship between variables: Covariance, Correlation, Regression(Linear, Logistic).

5. Probability Distribution Functions: Probability mass function, Probability density function, Cumulative density function.

6. Distribution type: Continuous distribution and discrete distribution.

7. Accuracy: True Positive, True Negative, False Positive and False-negative, Sensitivity, Specificity, Precision.

8. Hypothesis Testing: This is one of the important topics in statistics. Steps for finding out the hypothesis:

a. State the null and alternative hypotheses.

b. Determine whether the test is a two-tail test or a single tail test.

c. Calculate the statistical value and probability value.

d. Based on your result accept or reject the null hypothesis.

· DATA CLEANING and EXPLORATORY DATA ANALYSIS(EDA):

Data cleaning mainly consists of examining the raw figures and transforming it into a more usable form. Basically, it gets the statistics ready for the developed model. The figures received by the data scientists are rarely in the state that it can be directly used. In this process you need to tackle to main problems that are:

1. Outliers- These are data points that are outlying from the actual bulk of data. Like if you need to calculate the mean of 5,3,4,6,2 it comes out to be 4. But if u replace 5 with 20 then your dataset becomes- 20,3,4,6,2 and your mean becomes-7. So the value “20” is shifting the mean value to 7. So it is an outlier.

2. Missing Value.- It is basically a null value or garbage value.

After this process is done, statistics are ready for EDA. In EDA, we uncover patterns and relationships in data. This process is invaluable to the business and can provide insights that could be previously unknown relationships between features, other actionable phenomena, or potentially even that the goal of the modeling project cannot be achieved with the figures available.

· DATA VISUALISATION:

After you get your results you need to visualize your result in the form of graphs of animation as not everyone in the company is a data expert.

You need to explain to them in layman terms what information you have gathered from the figures. You can use various visualization software like:

1. Tableau.

2. Power BI.

3. Periscope Data.

4. Sisense.

5. IBM Cognos Analytics.

6. GoodData

7. Thought spot

8. Amazon Quicksite

· PROGRAMMING LANGUAGES:

Since the jobs are growing in this sector there are coding languages that every fresher must know to grab the opportunity. The demand for technologies like cloud, artificial reality (AR), virtual reality (VR), artificial intelligence (AI), machine learning (ML), and deep learning has increased in the market. These technologies require programming knowledge. Some of the programming languages used are:

1. Python.

2. R Programming.

3. Java.

4. JavaScript.

5. SAS(Statistical Analysis System)

6. Scala.

7. TensorFlow.

8. Hadoop.

9. C#.

10. Ruby.

· MACHINE LEARNING:

Machine learning is a hot topic when it comes to questions like how to become a data analyst or how to start a career in data science. With the number of statistics out in the market, we can gain new and deep knowledge about topics through machine learning.

Topics that you need to learn for a data scientist career path are:

1. Supervised Learning and Unsupervised Learning.- These are different types of how we can train a machine. Get acquainted with supervised and unsupervised learning algorithms like:

· Supervised Learning algorithms:

1. Linear Regression

2. Logistic Regression.

3. K-Nearest Neighbour(KNN).

4. Random Forest.

· Unsupervised Learning Algorithms:

1. Clustering.

2. Association.

2. Neural Network(NN)-It is more complex than traditional learning methods. It is derived from a biologically inspired paradigm, where the input parameters are provided to different neurons which help to create machine learning algorithms.

3. Deep-Neural Network.-Here various layers of neural networks are stacked together to create a larger network to get output.

4. Ensemble Learning.-Improves ML by combining various other ML models.

5. Gradient boosted decision tree-It is an ensemble learning technique where predictors(input parameters) are made sequential.

6. Overfitting and Underfitting.

7. Regularization.- This is a process of modifying the algorithm to avoid overfitting.

8. Cross-validation- used for evaluating the trained Machine learning model

These are the basic topics you need to get friendly with to kick start your data scientist career path.