100DaysOfMLCode

Day 1: Text preprocessing

• Text preprocessing is one of the important step while working on text.
• We have to spend more time in cleaning the data to get better results.
• I have created a module to preprocess text.
• Link to the module: Text preprocessing

Day 2: Pandas: Part 1

• Pandas is one of the most used libraries by Data scientists.
• It contains some basic usage of Pandas.
• Link to the notebook: Pandas Part 1
• Source - Data analysis in Python with pandas
• Link to the official Pandas Documentation

Day 3: Pandas: Part 2

• They are many tricks which makes our life easier instead of wasting time in writing long codes. So, I have explored Pandas library thoroughly so that I can save a lot of time while coding.
• Please feel free to explore the notebook.
• Link to the notebook: Pandas Part 2
• Source - Data analysis in Python with pandas
• Link to the official Pandas Documentation

Day 4 and Day 5:

• Participated in the ZS Challenge organised by Hacker Earth. It was a four day challenge.
• Link to the Challenge
• Link to the Repository
• The Hackathon was awesome and the problem was challenging enough to test your skills.
• Secured a rank of 223 out of 4743 (top 5%).

Day 6 - Day 12:

• Participated in the Machine Learning Challenge - "Predict the damage to a building" organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• It is a Classification problem which involved a lot of preprocessing, joining tables. The data is huge which also helps us to solve the challenge in an optimized way.
• I faced a lot of problems like Data balancing, duplicate data, hyper parameter tuning. It's quite interesting.
• This project helped me alot in learning different ways to handle huge data.

Day 13:

• Participated in the American Express AI Challenge(Problem 2) - "Supervised Modeling with Emphasis on LAUC " organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• It is a Binary Classification problem. This problem involves chosing the correct algorithm(modelling).

Day 14:

• Data visualization using Seaborn
• Link to the Notebook
• Link to the official Seaborn Documentation
• Source - Python for Data Visualization - using Seaborn
• This Notebook contains different plots that were explored on sample datasets, that were provided by the Seaborn library.

Day 15 - Day 19:

• Resumed my work on Machine Learning Challenge - "Predict the damage to a building" organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• Increased my score from 0.68571 to 0.72056 after spending a lot of time in tuning hyperparameters.

Day 20:

• A Complete Tutorial on Tree Based Modeling
• This tutorial explains about various topics such as:
  • What is a Decision Tree? How does it work?
  • Regression Trees vs Classification Trees
  • How does a tree decide where to split?
  • What are the key parameters of model building and how can we avoid over-fitting in decision trees?
  • Are tree based models better than linear models?
  • Working with Decision Trees in R and Python
  • What are the ensemble methods of trees based model?
  • What is Bagging? How does it work?
  • What is Random Forest ? How does it work?
  • What is Boosting ? How does it work?
  • Which is more powerful: GBM or Xgboost?
  • Working with GBM in R and Python
  • Working with Xgboost in R and Python
  • Where to Practice ?
• The tutorial is very intuitive and any one can understand the concepts easily. It is worth reading and spending time.

Day 21 - Day 23:

• Resumed my work on Machine Learning Challenge - "Predict the damage to a building" organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• Increased my score from 0.72056 to 0.72498 after spending a lot of time in tuning hyperparameters.

Day 24 - Day 25:

• Resumed my work on American Express AI Challenge(Problem 2) - "Supervised Modeling with Emphasis on LAUC " organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• Increased my score from 0.96655 to 0.966556 after spending a lot of time in feature extraction and tuning hyperparameters.

Day 26:

• Successfully completed the challenge - American Express AI Challenge(Problem 2) - "Supervised Modeling with Emphasis on LAUC" organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• Rank: 36 out of 1354 participants

Day 27 - Day 29:

• Successfully completed the Machine Learning Challenge - "Predict the damage to a building" organised by Hacker Earth.
• Link to the Challenge
• Link to the Repository
• Rank: 242 out of 7540 participants.

Day 30 - Day 35:

• Participated in the Machine Hack Challenge - "Predict house proces in Bangalore".
• Link to the Challenge
• Link to the Repository
• It is a Regression problem. It involves a lot preprocessing and handling missing values. It's really fun though.

Day 36 - Day 38:

• Started reading about Time series analysis from scratch.
• Text Book: Time Series Analysis Forecasting and control by Box and Jenkins.
• Completed reading first 2 chapters - This gives introduction to time series analysis and the types of methods used.

Day 39 - Day 40:

• Resumed my work on the Machine Hack Challenge - "Predict house proces in Bangalore".
• Link to the Challenge
• Link to the Repository

Day 41 - Day 44:

• Read third chapter from the text book: Time Series Analysis Forecasting and control by Box and Jenkins.
• Chapter 3: Linear Statioanry models (Autoregressive, Moving average, Mixed Autoregressive moving average). It also explains about the conditions of stationarity, invertibility, Autocorrelation function and Partial autocorrelation functions.
• This text book explains the concepts mathematically in a very simple way.

Day 45 - Present:

• Working on different Time Series data.
• Appying different stochastic models and comparing with latest models.