Data science lifecycle focus on machine learning and various analytical techniques to glean insights and forecasts from data in order to achieve business objectives.

Machine learning algorithms and statistical techniques are used in a comprehensive science lifecycle process to produce improved prediction models. The process involves numerous common data science steps, including data extraction, preparation, cleaning, modeling, and evaluation, among others.

The life cycle of data science involves multiple steps.

1. Data Acquisition

2..Data Processing

3. Model Building

4. Pattern Evaluation

5. Knowledge Representation

Any type of model involves two types of variables, independent and dependent which would be defined as X and Y respectively. X can hold any number of independent variables, whereas Y holds only one value which is actual or predictable (assumption).

Y = Dependent Variable

X = Independent Variable

A machine learning model identifies the relation between Y and X.

Based on data, we need to predict the customer churn. What could be the primary drivers in terms of customer churn?

If the dependent variable is churn, then x would be the value that helps in predicting the dependent variable.

For example, if Y = Churn of data which affects the business

then X1 = Price, X2= Customer's demographic, X3 = network quality, X4 = tenure of the customer, X4 = service and X5= complaints 

We may also outline the data which is not required for our data modeling or prediction. The process of eliminating the poorly measured data is called data outliers.

You should discard the outlier if it is clear that the error resulted from poorly measured or inputted data. You can exclude the outlier if it has no effect on the results but does influence the assumptions. But make a note of that in a footnote in your essay.

R Language Data types help you to define, calculate and eliminate unwanted data. Especially the definition stage involves different inputs and those inputs would be defined using various data types.

R - Language - Vector is a Homogeneous data type that holds the same type of data inputs.

List - Heterogeneous - information should be in the same column of an excel file. The information can be heterogeneous 

What is homogeneous data?

If a data set contains items that are comparable to one another, it is homogenous. In the context of this article, it refers to information that came from the same source. In a typical supervised learning scenario, this will lead to the data set having the exact same label applied to it throughout.

What is heterogeneous?

Any data having a large degree of variation in data kinds and formats is considered heterogeneous. As a result of missing values, high data redundancy, and untruthfulness, they may be unclear and of low quality. To address the needs for business information, diverse data requires tough integration.

What is one-dimensional?

One dimension is used to partition the application data in single-dimensional data categorization. When you need to divide application data into discrete groups, such as those based on years or quarters, use a single-dimensional data categorization. Time is a common dimension in single-dimensional data classifications.

what is multi-dimensional?

A technique for organizing and properly assembling the contents of the database, the multi-Dimensional Data Model is used to sort data in the database. Contrary to relational databases, which allow users to access data through queries, the multidimensional data model enables users to query analytical questions related to market or business trends. They enable consumers to quickly acquire responses to their inquiries by producing and analyzing the data very quickly.

---

Also Read:

Deep Learning of Data Science To Get Started

Logistic Regression Model using Python

Supervised and Unsupervised Learning

Introduction to Data Science

Share this course on your social media to help your peers. Follow our Data Science page on WEBYPOST for more courses on data science