undefined. & Genetics. PRESENTED BY. Kartik 2101730051. 

Scene 1

[Audio] An overview of our project is provided in this slide. It starts with a timeline outline for the project followed by data on the datasets to be used. Further details on the data preprocessing and feature engineering techniques used to prepare the data for the Machine Learning model will be provided. Lastly the ML Model will be presented along with the accuracy result and expected outcome.. 

Scene 2

[Audio] The goal of this project is to create a machine learning model to predict genes associated with genetic disorders and classify genetic disorders for accurate diagnosis. We will collaborate with geneticists and medical professionals and utilize curated datasets while employing cutting-edge feature engineering tactics to identify significant genetic characteristics. Our objective is to enable early detection and personalized treatment for the public promote precision medicine and scientific research in genetics and guarantee model accuracy and applicability.. 

Scene 3

Develop a machine learning model for predicting genes associated with genetic disorders.    Train the model to predict the likelihood of genetic disorders based on input genetic data.
 Implement subclass classification to categorize genetic disorders for precise diagnosis.
 Utilize curated genetic datasets for model training and validation.
 Apply advanced feature engineering techniques to extract meaningful genetic features.
 Collaborate with geneticists and medical professionals to ensure model accuracy and relevance.
 Aim to facilitate early detection and personalized treatment of genetic disorders.
 Contribute to advancements in precision medicine and genetic research.

[Audio] The project has been broken down into three distinct stages. Initially we chose our idea assessing the problem. Subsequently we concentrated on data pre-processing and attribute extraction. Finally we implemented the model itself. Further technical details of each of the stages will be provided in the subsequent slides.. 

Scene 4

IDEA SELECTION AND UNDERSTANDING THE PROBLEM

DATA PRE PROCESSING AND FEATURE EXTRACTION

MODEL IMPLEMENTATION AND ACCURACY OPTIMIZATION

[Audio] The dataset for this project has 22083 records for training and 9465 for testing. Each record contains 43 attributes plus a target variable denoting the Genetic disorder and its subclass. The machine learning model will be deployed to predict the target variable based on the other fields in the dataset.. 

Scene 5

The dataset contains the following files: train.csv: Training Data (22083 x 45) test.csv: Testing Data (9465 x 43)
 The following target variables are included in the dataset:  Genetic disorder Disorder subclass
 The machine will learn to predict the target variable based on 43 other fields in addition to the target variable.

[Audio] We explored our dataset to understand the relationships with genetic disorders and identified any key statistical summaries or correlations that would be helpful in feature selection. We addressed any missing values and managed any outliers present. Visualizations were used to better interpret and identify any useful patterns in our data and then we optimized the feature set by removing any irrelevant or redundant features that were identified through our exploratory data analysis. This powerpoint slide describes the data preprocessing step we took while developing our machine learning model.. 

Scene 6

Explored dataset features to understand distributions and relationships with genetic disorder.
 Identified key statistical summaries and correlations to guide feature selection.
 Addressed missing data to ensure dataset completeness.
 Detected and managed outliers using visualization techniques and statistical methods.
 Utilized visualizations to interpret data patterns and aid in feature selection.
 Optimized feature set by removing irrelevant or redundant features based on EDA insights.

[Audio] We are utilizing four different algorithms to predict accurate results for this Machine Learning Model CatBoost Random Forest XGBoost and LightGBM. CatBoost is crafted to produce robust performance with automatic feature scaling and effective handling of categorical features. Random Forest aids in offering strong performance through ensemble learning and further grants feature importance analysis and interpretability. XGBoost is used to achieve high accuracy and efficiency with optimized gradient boosting algorithms and hyperparameter tuning. Lastly LightGBM supplies efficient training and prediction speeds using histogram based algorithms for greater accuracy and scalability.. 

Scene 7

1. CatBoost: Achieved robust performance with automatic feature scaling and effective handling of categorical features.
 2. RandomForest:  Demonstrated strong performance through ensemble learning and provided good interpretability with feature importance analysis.
 3. XGBoost:  Delivered high accuracy and efficiency with optimized gradient boosting algorithms and enhanced performance through hyperparameter tuning.
 4. LightGBM:  Showcased efficient training and prediction speeds using histogram-based algorithms for improved accuracy and scalability.

[Audio] Our model with the four machine learning algorithms CatBoost RandomForest XGBoost and LightGBM has achieved remarkable accuracy rates. We obtained an overall accuracy of 56% with CatBoost 57% with RandomForest 59% with XGBoost and 56% with LightGBM. We have seen even more impressive results for subclass prediction accuracy 68% with CatBoost 70% with RandomForest 71% with XGBoost and 70% with LightGBM. We are extremely pleased with these results and extend our sincere gratitude to the team for their hard work and commitment.. 

Scene 8

Overall Accuracy:
 CatBoost: Achieved an accuracy of 56%. RandomForest: Attained an accuracy of 57%. XGBoost: Demonstrated an accuracy of 59%. LightGBM: Showcased an accuracy of 56%.
 Subclass Prediction Accuracy:
 CatBoost: Achieved a subclass prediction accuracy of 68%. RandomForest: Attained a subclass prediction accuracy of 70%. XGBoost: Demonstrated a subclass prediction accuracy of 71%. LightGBM: Showcased a subclass prediction accuracy of 70%.

Black Orange Bold Modern Project Proposal Presentation

Scene 0

An overview of our project is provided in this slide. It starts with a timeline outline for the project
 followed by data on the datasets to be used. Further details on the data preprocessing and feature engineering techniques used to prepare the data for the Machine Learning model will be provided. Lastly
 the ML Model will be presented
 along with the accuracy result and expected outcome.

avatar

The goal of this project is to create a machine learning model to predict genes associated with genetic disorders and classify genetic disorders for accurate diagnosis. We will collaborate with geneticists and medical professionals
 and utilize curated datasets while employing cutting-edge feature engineering tactics to identify significant genetic characteristics. Our objective is to enable early detection and personalized treatment for the public
 promote precision medicine and scientific research in genetics
 and guarantee model accuracy and applicability.

The project has been broken down into three distinct stages. Initially
 we chose our idea
 assessing the problem. Subsequently
 we concentrated on data pre-processing and attribute extraction. Finally
 we implemented the model itself. Further technical details of each of the stages will be provided in the subsequent slides.

The dataset for this project has 22083 records for training and 9465 for testing. Each record contains 43 attributes
 plus a target variable denoting the Genetic disorder and its subclass. The machine learning model will be deployed to predict the target variable based on the other fields in the dataset.

We explored our dataset to understand the relationships with genetic disorders and identified any key statistical summaries or correlations that would be helpful in feature selection. We addressed any missing values and managed any outliers present. Visualizations were used to better interpret and identify any useful patterns in our data
 and then we optimized the feature set by removing any irrelevant or redundant features that were identified through our exploratory data analysis. This powerpoint slide describes the data preprocessing step we took while developing our machine learning model.

We are utilizing four different algorithms to predict accurate results for this Machine Learning Model CatBoost
 Random Forest
 XGBoost
 and LightGBM. CatBoost is crafted to produce robust performance with automatic feature scaling
 and effective handling of categorical features. Random Forest aids in offering strong performance through ensemble learning and further grants feature importance analysis and interpretability. XGBoost is used to achieve high accuracy and efficiency with optimized gradient boosting algorithms and hyperparameter tuning. Lastly
 LightGBM supplies efficient training and prediction speeds using histogram based algorithms for greater accuracy and scalability.

Our model with the four machine learning algorithms  CatBoost
 RandomForest
 XGBoost
 and LightGBM  has achieved remarkable accuracy rates. We obtained an overall accuracy of 56% with CatBoost
 57% with RandomForest
 59% with XGBoost
 and 56% with LightGBM. We have seen even more impressive results for subclass prediction accuracy  68% with CatBoost
 70% with RandomForest
 71% with XGBoost
 and 70% with LightGBM. We are extremely pleased with these results and extend our sincere gratitude to the team for their hard work and commitment.