SRES’

SANJIVANI COLLEGE OF ENGINEERING,

KOPARGAON
423601(M.S.)

Department of
Mechatronics Engineering 2022-2023
REPORT ON

“Salary Prediction Based on work

experience ML Web App”

1|Page
 CERTIFICATE

This is to certify that this report on Project entitled, “Salary Prediction Based on
work experience ML Web App”
Submitted by,

Sr. No Roll no. NAME PRN NO

1 28 Kale Kalyan Jayant UMX20M1028

For the partial fulfilment of the requirements of Final Year (Mechatronics Engineering) degree
of the Sanjivani COE, Kopargaon embodies the work done by them under our guidance and
supervision in the academic year 2023-2024

Prof. Chaitanya Kale

(Subject-ML)

Prof. R. A. Kapgate Dr. A. G. Thakur

Head of Dept. of Mechatronics Engg. Director, SCOE

2|Page
 DECLARATION

We declare that this written submission represents our ideas in our own words; we have
adequately cited and referenced the original sources. We also declare that we have adhered to all
principles of academic honesty and integrity and have not misrepresented or fabricated any
idea/data/fact/source in our submission. We understand that any violation of the above may
cause disciplinary action by the Institute and can also evoke penal action from the sources which
have not been properly cited.

Date: - 18 /05/2023

Place: - Kopargaon

Sr no. Roll no. Name

1 28 Kale Kalyan Jayant

3|Page
 ACKNOWLEDGEMENT

"Salary Prediction Based on work experience ML Web App ” has been the

opportunity to express ourselves technically. This has proven to be a steppingstone which will be

of immense help to us as we enter market. We want to express our gratitude to everyone who

helped us by giving moral support and by solving our difficulties. Everyone has contributed

immensely and helped us for the same into the completion of project.

We take this opportunity to express our deep sense of gratitude towards head of

department of Mechatronics Engineering Dr. R. A. Kapgate and our esteemed guide Prof.

Chaitanya Kale for his expert guidance during preparation of this seminar. He has received us

whenever we required his help. In true sense of word, we are grateful to him. We are highly

grateful to our subject coordinator for extending all the facilities in completing this seminar.

We would also like to thank all our friends, who helped us and initiated discussion during

the seminar. Last but not least; we want to acknowledge our beloved parents, who have taken

great pains for our education.

Signature
28 Kale Kalyan Jayant

4|Page
 ABSTRACT
This report presents a comprehensive overview of a Machine Learning (ML) web application
designed to predict salaries based on an individual's years of experience and job type. The
primary objective of this project is to harness the power of data transformation and machine
learning algorithms to develop a predictive model that offers valuable insights into salary
expectations.The dataset employed in this study is characterized by its simplicity, with minimal
missing data, and comprises a training dataset featuring key features such as years of experience
and corresponding salary information. Through rigorous data transformation processes, the raw
data is refined to enhance the model's accuracy and reliability.The predictive power of the model
hinges on two fundamental factors: years of experience and job type. By analyzing these key
parameters, the ML model provides a reliable guide for estimating salaries, making it a valuable
tool for both employers and employees in the decision-making process related to
compensation.This web application aims to serve as a practical resource for organizations and
individuals seeking informed salary predictions. The simplicity of the model, combined with its
demonstrated accuracy in predicting salaries based on years of experience, positions it as a
valuable asset for shaping compensation strategies. Ultimately, this project contributes to the
ongoing dialogue surrounding fair and data-driven salary determination, offering a useful tool
for stakeholders navigating the intricacies of compensation decisions in various professional
settings.

5|Page
 INTRODUCTION

OBJECTIVES
 Develop a robust Machine Learning (ML) web application with the primary goal of
predicting salaries based on years of experience and job type.
 Implement advanced data transformation techniques to preprocess and refine the raw
dataset, ensuring minimal missing data and enhancing the accuracy of the predictive
model.
 Leverage machine learning algorithms to create a predictive model that provides valuable
insights into salary expectations, focusing on key parameters such as years of experience
and job type.
 Demonstrate the practical utility of the web application as a resource for organizations
and individuals by offering reliable and reasonable predictions for salary determination.
 Contribute to the advancement of fair and data-driven salary determination practices by
developing a model that serves as a guide for compensation decisions, fostering informed
dialogue in professional settings.
 Provide a user-friendly and accessible platform that can be utilized by both employers
and employees, facilitating better decision-making processes related to compensation and
contributing to a more transparent salary negotiation environment.

LIMITATIONS:
 Limited Feature Set: The predictive model relies primarily on years of experience and job
type, potentially overlooking other relevant factors that could influence salary, such as
education level, geographic location, or industry-specific qualifications.
 Data Quality Constraints: The accuracy of the model is contingent upon the quality of the
training dataset. If the dataset contains inaccuracies, biases, or insufficient representation
of diverse job types, the model's predictions may be compromised.
 Static Model: The web application assumes a static relationship between years of
experience, job type, and salary. It does not account for dynamic changes in job markets,
economic conditions, or emerging industry trends, potentially limiting its adaptability
over time.
 Assumption of Linearity: The model assumes a linear relationship between the selected
features and salaries. Non-linear relationships or interactions between variables may not
be adequately captured, affecting the accuracy of predictions.
 Generalization Challenges: The model's performance may vary when applied to different
industries, job markets, or regions, as it may not account for the nuanced salary
determinants specific to each context.

6|Page
  Lack of Individual Context: The model treats individuals as homogeneous entities,
neglecting personal attributes and circumstances that could significantly impact salary
expectations, such as negotiation skills, individual achievements, or unique job
responsibilities.

REQUIREMENTS SPECIFICATION

 Data Analysis and Visualization:

The system must support data analysis methods to explore and understand the
dataset.Users should be able to visualize data trends, distributions, and relationships between
variables.Utilize Seaborn, Matplotlib, and other relevant libraries for effective data
visualization.

 Linear Regression:
Implement linear regression algorithms for modeling the relationship between years
of experience, job type, and salary.Ensure the model is capable of making accurate
predictions based on linear relationships.

 Polynomial Transformation:
Apply polynomial transformation techniques to capture non-linear relationships
between features and salaries.The system should allow users to choose polynomial degrees
for optimization.

One litre of water was poured into

the electrical kettle.
2. The temperature of the water is
determined by using a
thermocouple.

7|Page
3. The kettle is placed on a
weighing scale and the kettle with
water included is weighed.
4. Switched on the electricity.
5. The temperature and the time
until the water is recorded.
6. The time and the reducing
weight are recorded as the water
evaporated.
7. The voltage and the current are
determined in order to determine
the electrical assumption for
the heater.
8. A table is constructed to obtain
the necessary data and the
following output is determined:

8|Page
I. Rate of temperature rise to boil
the water (Table and graph of
Temp vs Time).
II. Rate of water evaporated from
the kettle (Table and graph Weight
of evaporated water
vs Time).
III. Rate of heat transferred of
water to boil.
IV. Rate of heat transferred of 0.5
litres of water to evaporate.
V. Total power consumption.
VI. The voltage and current from
the power supply are determined to
determine power
input.
VII. Efficiency of the kettle.
9|Page
One litre of water was poured into
the electrical kettle.
2. The temperature of the water is
determined by using a
thermocouple.
3. The kettle is placed on a
weighing scale and the kettle with
water included is weighed.
4. Switched on the electricity.
5. The temperature and the time
until the water is recorded.
6. The time and the reducing
weight are recorded as the water
evaporated.
7. The voltage and the current are
determined in order to determine
the electrical assumption for
10 | P a g e
the heater.
8. A table is constructed to obtain
the necessary data and the
following output is determined:
I. Rate of temperature rise to boil
the water (Table and graph of
Temp vs Time).
II. Rate of water evaporated from
the kettle (Table and graph Weight
of evaporated water
vs Time).
III. Rate of heat transferred of
water to boil.
IV. Rate of heat transferred of 0.5
litres of water to evaporate.
V. Total power consumption.

11 | P a g e
VI. The voltage and current from
the power supply are determined to
determine power
input.
VII. Efficiency of the kettle.
One litre of water was poured into
the electrical kettle.
2. The temperature of the water is
determined by using a
thermocouple.
3. The kettle is placed on a
weighing scale and the kettle with
water included is weighed.
4. Switched on the electricity.
5. The temperature and the time
until the water is recorded.

12 | P a g e
6. The time and the reducing
weight are recorded as the water
evaporated.
7. The voltage and the current are
determined in order to determine
the electrical assumption for
the heater.
8. A table is constructed to obtain
the necessary data and the
following output is determined:
I. Rate of temperature rise to boil
the water (Table and graph of
Temp vs Time).
II. Rate of water evaporated from
the kettle (Table and graph Weight
of evaporated water
vs Time).
13 | P a g e
III. Rate of heat transferred of
water to boil.
IV. Rate of heat transferred of 0.5
litres of water to evaporate.
V. Total power consumption.
VI. The voltage and current from
the power supply are determined to
determine power
input.
VII. Efficiency of the kettle.
Technologies/Libraries Used:

 Develop the web app using Python 3 as the primary programming language.

 Utilize Pandas for efficient data manipulation and preprocessing.

 Incorporate NumPy for numerical operations and array manipulations.

 Employ Scikit-learn for machine learning model implementation.

 Integrate Jupyter notebooks for interactive development and documentation.

14 | P a g e
1. Data Overview:

The dataset employed for this salary prediction model is characterized by its simplicity, with

minimal missing data.Raw data includes a training dataset containing features such as years of

experience and their corresponding salaries.A subset of 20% of the training dataset was extracted

to form a test dataset for evaluating the model's performance.

2. Testing Dataset:

A separate testing dataset was utilized for simulating real-world scenarios. This dataset lacks

salary information, resembling situations where predictions are needed for new, unseen data.

3. Information Used for Prediction:

15 | P a g e
The primary feature used for predicting salaries is the 'Years of Experience.'

4. Model Training:

The model training process is encapsulated in the 'model.py' script.Upon completion of training,

the model is saved as 'model.pkl,' a pickle file, for future use and deployment.

5. Run App:

The 'app.py' script is responsible for managing the Flask application and handling APIs.

To initiate the web application, users need to open the command prompt, navigate to the

specified directory, and execute 'python app.py.'

16 | P a g e
17 | P a g e
6. Procedure:

The following steps outline the procedure to run the web application:

Open the command prompt.

Navigate to the designated directory.

Execute the command 'python app.py.'

7. Model Files:

'model.pkl': The serialized pickle file containing the trained machine learning model.

'model.py': Script responsible for training the model and saving it as 'model.pkl.'

8. Project Objective:

The primary goal of this project is to predict employee salaries based on their years of

experience.

9. Model Output:

The trained model provides salary predictions based on the input of years of experience, offering

a valuable tool for compensation decisions.

18 | P a g e
10. Real-world Application:

The model's use of a testing dataset without salary information mirrors real-world scenarios

where predictions are needed for new data points.

11. Flask Application:

The Flask application, managed by 'app.py,' serves as the user interface for interacting with the

model predictions.

19 | P a g e
 Conclusions

In conclusion, the provided data and model for the Salary Prediction Web App present a
straightforward yet effective approach to forecasting employee salaries based on years of
experience. The dataset's simplicity, with minimal missing data, facilitates model training and
evaluation. The inclusion of a testing dataset without salary information mimics real-world
scenarios, enhancing the model's applicability to new and unseen data.

The model training process, encapsulated in 'model.py,' yields a serialized pickle file
('model.pkl') for easy deployment and future use. Running the web application, managed by
'app.py' with Flask, is a user-friendly process, simplifying the prediction of salaries through the
input of years of experience.

This project successfully achieves its primary objective of predicting salaries based on a single
feature, providing a valuable tool for compensation decisions. The combination of machine
learning techniques and a Flask application offers a practical solution for organizations and
individuals seeking informed salary predictions.

Looking forward, potential enhancements could include expanding the feature set for more
comprehensive predictions and refining the user interface for an enhanced user experience.
Overall, this project lays a solid foundation for the integration of data science and web
development to address the practical challenges of salary prediction in professional settings.

20 | P a g e