Light

zacklarsen / kaggle-wids-datathon-2020 Goto Github PK

View Code? Open in Web Editor NEW

0.0 0.0 0.0 2.93 MB

Kaggle competition for survival analysis, sponsored by Stanford

Jupyter Notebook 99.85% Python 0.15%

kaggle-wids-datathon-2020's Introduction

Hi there 👋

🔭 I’m a data scientist interested in clinical predictive models, clinical decision support tools, causal inference, MLOps tooling & machine learning engineering, deep learning, developer productivity tools and approaches, optimization problems, interactive data visualization, analytics and data engineering concepts, and product management.

My github pages sites:

My Repos:

kaggle-wids-datathon-2020's People

Watchers

kaggle-wids-datathon-2020's Issues

Create script for multiple data imputation

Install hydra and create first configuration

Use summaryTools for EDA

Replicate this model approach

https://www.kaggle.com/code/danofer/wids-2020-starter-catboost-0-9045-lb/notebook

Define sklearn preprocessing pipeline

Add Prefect parameters with Hydra config

Track artifacts with MLflow

Adapt scikit-survival to WiDS

Add more metrics to evaluate.py step

Setup Prefect

Create Kaplan-Meier plots

Make Quarto EDA notebook

Add sklearn transformer pipeline

Random Survival Forest model trained

Implement Random Survival Forests using the scikit-survival package. Experiment with different hyperparameters to improve the accuracy metrics and performance.

Create script for DeepSurv model

Create EDA notebook

Kaplan-Meier model trained

Use prefect-viz to display flow

Add Featuretools feature engineering notebook

Add Mermaid workflow diagram

Make script for Cox proportional hazards model

Train automl model with mlflow and maybe PyCaret

Add encounter_id from held-out data to preds somehow to ensure correct encounter_id is in formatted predictions/submission

Test parameters in Prefect

Parameters seem to be simple keyword args which I already had. Hydra is better for supplying these values in a config anyway.

Impute with IterativeImputer

https://scikit-learn.org/stable/modules/generated/sklearn.impute.IterativeImputer.html#sklearn.impute.IterativeImputer

Perform feature engineering with featuretools

DeepSurv model trained

Explore the DeepSurv architecture for survival analysis. Understand how it extends the Cox model and evaluate its performance on the dataset.

Format final predictions

Pandas Profiling

Make late submission

Use Yellowbrick or explainer dashboard to visualize model performance

Try Prefect UI

Use ydata-profiling on data

Cox Proportional Hazards model trained

Explore the implementation of Cox Proportional Hazards Model using the lifelines library. Evaluate the model's performance and interpret the results.

Complete data prep work

Generate predictions

Time-Varying Cox regression model trained

Extend the Cox model to handle time-varying covariates. Pre-process the dataset into a "long" format and evaluate the model's performance.

Create virtual environment and requirements.txt file

Add Pygwalker notebook for EDA

Add issues for each approach in this post (see notes)

Train benchmark catboost model

Set up MLflow recipes layout

Define and test format.py step

Track artifacts with Prefect artifacts

https://docs.prefect.io/latest/concepts/artifacts/

Install Dagster and Dagit

Install Anaconda properly on local computer and create conda env with pandas_profiling

Read this article

Integrate MLflow with training runs

Train logistic regression model with Hydra config and Prefect

Add Prefect flow from notebook to script

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.