- Project Overview
- Business Understanding
- Objectives
- Data Understanding
- Exploratory Data Analysis (EDA)
- Modeling
- Recommendations
- Conclusion
- Next Steps
The Movie Rating Prediction project involves analyzing a dataset containing information about Indian movies. The dataset includes details like movie name, year, duration, genre, rating, votes, director, and three main actors. The data will be used to build a predictive model for movie ratings and extract valuable insights from the movie industry.
The film industry relies on understanding the factors that influence movie success. Accurately predicting movie ratings can aid in decision-making, such as choosing the right actors, directors, and genres, as well as determining marketing strategies
- Develop a predictive model: Create a machine learning model to predict movie ratings based on the provided dataset.This is essentially a regression problem, where we aim to estimate the numerical movie ratings based on various features.
- Identify influential factors: Analyze the dataset to determine which factors (e.g., genre, director, actors) have the most significant impact on movie ratings.
- Provide actionable insights: Offer insights to the film industry stakeholders to make informed decisions about movie production, casting, and marketing.
The dataset is obtained from Kaggle: IMDb India Movies
The dataset has the following columns:
Name: Movie name
Year: release year of the movies
Duration: Movie duration
Genre: Movie genre
Rating: Movie rating
Votes: Number of votes received
Director: Movie director
Actor 1: First main actor
Actor 2: Second main actor
Actor 3: Third main actor
EDA inlude Rating analysis, Top 10 Directors with the Most Movies Directed, Top 10 Actors with the Most Movie Appearances, Top 10 Directors with the Highest-Rated Movies, Top 10 Highly Rated Movie Genres, Top 10 Years with Highest Average Ratings and Duration vs Rating
- Baseline Model - Linear Regression
- Second Model: Random Forest Model
- Third Model: Gradient Boosting Regressor
- Hyperparameter Tuning of Random Forest Model
The tuned Random Forest performed better than the tuned model in terms of model generalization and avoiding overfitting, due to its more balanced performance between training and test data. This will be the final model used for prediction.
Features that had the most significant impact on the target variable in the above model:
- Year
- Votes
- Duration
- Year: Stay current with trends for successful movie releases.
- Votes: Encourage audience reviews for more reliable ratings.
- Duration: Align with audience preferences for movie length.
- Collaboration: Partner with established industry figures.
- Genres: Invest in highly-rated genres like History and Romance.
- Duration vs. Ratings: No strong duration-rating correlation.
- Critical Acclaim: Aim for positive reviews and acclaim
In conclusion, this project provides valuable insights and a predictive model for movie rating prediction. The film industry can benefit from these findings to make data-driven decisions regarding movie production, casting, and marketing. The most influential factors identified are the year of release, the number of votes, and movie duration.
- Model Refinement: Improve the rating prediction model with advanced machine learning methods.
- Feature Engineering: Enhance the model's performance by experimenting with new features.
- User Reviews Analysis: Analyze user reviews and sentiments for deeper audience insights.
- Real-time Data: Implement real-time data updates to keep the model current with the latest trends