INTRODUCTION TO DATA SCIENCE

ASSIGNMENT NO.1

Submitted To: Dr. Abid Sohail

Submitted by: Shifa Asif (F2020331011)

 1. What are monitoring reports?

Monitoring reports in data science provide regular updates on system performance and
behavior, typically generated by monitoring tools tracking specific metrics or KPIs.
Monitoring reports provide data scientists with a comprehensive understanding of their
data science system, models, or algorithms, identifying anomalies and facilitating
evaluation, improvements, and optimizations to ensure data quality, model performance,
and reliability.
Monitoring reports include metrics like data availability, quality, model accuracy, and
performance, generated regularly and often with visualizations, graphs, or charts for
easy interpretation and decision-making.

2. What are analytical reports?

Analytics reports discuss the analysis and insights obtained from data through a variety
of analytical methods. These reports offer useful facts and practical suggestions derived
from data-driven decision making to stakeholders. Numerous sectors, including
marketing, operations, finance, and consumer behavior, can produce analytics reports.

In analytics reports, patterns, trends, and linkages are often found through data
collection and analysis. Statistical modelling, machine learning, predictive analytics, and
other cutting-edge methods are used in this investigation. Analytics reports can handle
and analyze data effectively by using a variety of tools and technologies.

Some common analytical reports are:

 Descriptive analytical report
 Diagnostic analytical report
 Predictive analytical report
 Prescriptive analytical report

3. What are daily reports?

As the name implies, daily reports are produced every day and offer a glimpse into the
different operations taking place within an organization. These reports, which can be
customized for particular divisions or roles within an organization, aid in performance
monitoring, bottleneck identification, and required corrections.
Important metrics and KPIs pertaining to a specific department or process are usually
covered in daily reports. For instance, in sales, daily reports can contain information on
revenue, conversion rates, and the quantity of leads generated. Daily reports in
manufacturing may cover equipment utilization, quality parameters, and production
output. Managers and supervisors frequently utilize these reports to pinpoint areas in
need of development and make choices in real time. Daily reports can be produced
manually or automatically by systems that gather and aggregate data.

4. Explain the SQL function (average, sum count) in details and also write the
importance of these functions in analytics reports.

Average: The average function in SQL is used to calculate the average value of a set of
numeric data. It takes a column or expression as input and returns the average value of
all the records in that column or expression.

Syntax: AVG (column_name)

Example: SELECT AVG (salary) FROM employees;

Importance in analytics reports: The average function is useful in analytics reports as

it helps to understand the central tendency of the data. It provides insights into the
typical value or level of a certain attribute. For example, in a sales report, the average
sales revenue per day can help identify the average performance or profitability.
Average is often used in combination with other functions such as grouping, filtering, or
aggregation to gain more detailed insights from the data.

Sum: The sum function in SQL is used to calculate the total sum of a set of numeric
data. It takes a column or expression as input and returns the sum of all the records in
that column or expression.

Syntax: SUM (column_name)

Example: SELECT SUM (quantity) FROM orders;

Importance in analytics reports: The sum function is essential in analytics reports for
aggregating and totaling numerical data. It helps understand the total value or quantity
of a certain attribute. For example, in a financial report, the sum of sales revenue or
expenses can provide insights into the overall performance of a business. Sum is
frequently used with grouping or filtering to analyze subsets of data or compare different
categories.
Count: The count function in SQL is used to calculate the number of records in a table
or the number of occurrences of a particular attribute in a table. It can be used with the
asterisk (*) wildcard to count all records or with a specific column name to count the
non-null values.

Syntax: COUNT (column_name) or COUNT (*)

Example 1: SELECT COUNT (*) FROM customers;

Example 2: SELECT COUNT (city) FROM customers;

Importance in analytics reports: The count function plays a crucial role in analytics
reports as it provides insights into the size, distribution, or prevalence of certain
attributes. It helps answer questions like "How many customers do we have?", "How
many orders were placed?", or "How many products were sold?". Count is often used in
combination with grouping, filtering, or joins to analyze subsets of data or compare
different categories. It helps in making data-driven decisions based on the volume or
presence of particular elements.

5. How the process of data science works using ML?

The process of data science typically involves the use of machine learning (ML)
algorithms. Here is a general overview of how the process works:

1) Problem Definition: Clearly define the problem or question you want to address
using data science and ML techniques. This could be anything from predicting
customer churn to classifying images.

2) Data Collection: Collect relevant data that will help you solve the problem. This
can include structured data (e.g., in spreadsheets or databases) or unstructured
data (e.g., text, images, audio).

3) Data Preprocessing: Clean and prepare the collected data. This involves
handling missing values, removing noise, scaling features, transforming
variables, and preparing the data in a format suitable for ML algorithms.

4) Exploratory Data Analysis (EDA): Understand the characteristics and

relationships within the data through visualization and statistical analysis. EDA
helps in identifying patterns, outliers, correlations, and other useful insights.

5) Feature Engineering: Select, create, or transform features from the available

data that are relevant to the problem. This step aims to improve the algorithm's
performance by providing the most informative and discriminative features.
6) Model Selection: Choose an appropriate ML model based on the problem type,
available data, and desired outcome. This can include decision trees, support
vector machines, neural networks, or ensemble methods.

7) Model Training: Split the dataset into training and validation sets. Train the ML
model using the training set, optimizing the model's parameters or
hyperparameters to minimize the error.

8) Model Evaluation: Evaluate the trained model's performance using appropriate

evaluation metrics, such as accuracy, precision, recall, or F1 score. Assess if the
model is satisfactory for deployment or if further iterations are required.

9) Model Deployment: Integrate the ML model into the desired system or

application to make predictions on new, unseen data. This could involve
deploying the model on edge devices, web servers, or cloud platforms.

10) Monitoring and Maintenance: Continuously monitor the performance of the

deployed model, retraining or fine-tuning it periodically to adapt to changing data
patterns or business requirements.

6. Draw a diagram to show the process of data analytics?

Define the
problem

Collect the data

needed for your
problem

Process the
data for
analysis

Explore the
data

Analyze the
data

Interpret results