#100DaysOfCode Day 100:
Finished Intermediate ML covering the final chapter on data leakage, and finished my 100 days!๐ฅณ
I have learnt about Python for data analysis, particularly financial time-series data, as well as the basics of SQL and made a start with machine learning!
#100DaysOfCode Day 99:
Completed the penultimate chapter for @kaggle's Intermediate Machine Learning course on XGBoost.
Learned how to use Extreme Gradient Boosting model and tweak the parameters such as n_estimators, early_stopping_rounds and learning_rate.
#100DaysOfCode Day 98:
Worked through the next chapter of @kaggle's intermediate ML course.
Covered the topic of cross-validation, to address some of the issues associated with randomly selecting validation data from small datasets.
#100DaysOfCode Day 97:
Making progress through @kaggle's Intermediate Machine Learning course.
Today learnt about using Pipelines to combine pre-processing and modelling steps to make for cleaner code.
#100DaysOfCode Day 96:
Finished the chapter on Categorical Variables for my intermediate ML course.
Learned how to combine multiple methods of dealing with categorical variables including testing cardinality and conditionally implementing label or one-hot encoding.
#100DaysOfCode Day 95:
Continued working through @kaggle's Intermediate Machine Learning micro-course, focusing on some methods of handling categorical variables such as Label Encoding and One-Hot Encoding.
#100DaysOfCode Day 94:
Began working through another of @kaggle's micro-courses: Intermediate ML.
So far I learnt how to manage different common issues in machine learning such as dealing with missing data values and handling categorical variables.
#100DaysOfCode Day 93:
Finished @kaggle's micro-course on intro to ML, and have been reading some research papers on ML uses in bioinformatics, particularly protein science, in preparation for my computational biology research internship.
#100DaysOfCode Day 92:
Made a start on improving my machine learning knowledge by working through some of the micro-courses on @kaggle.
So far it's been explained super intuitively with practical exercises at every step!
#100DaysOfCode Day 90:
Implemented quantopian's open source Pyfolio library to further analyse algorithm backtests.
The built-in functionality of Pyfolio such as 'tear sheet' reports and plots are super useful in the assessment of an algorithm's performance.
#100DaysOfCode Day 89:
Continued working through the final chapter of my python course, Advanced Quantopian and Trading Algorithms.
Should finish the course in the next few days ๐
#100DaysOfCode Day 88:
Learned about hedging to minimise the effect of beta and essentially limit returns to alpha.
This greatly reduces volatility but also reduces expected return by limiting upside potential.
#100DaysOfCode Day 87:
Today I transferred the knowledge on creating pipelines in the @quantopian notebooks into implementing a pipeline within the IDE for use in a trading algorithm.
#100DaysOfCode Day 86:
Learned about creating pipelines in quantopian notebooks for research purposes, including implementing factors, classifiers, filters and screens.