ML - The way the world works - analyzing how things work
David Nishimoto
Machine learning is the most important technological breakthrough in the 21st century. Listen to my views on the future of machine learning. Code demonstrations on YouTube under my channel David Nishimoto
Episodes
Mentioned books
May 7, 2020 • 11min
Kmeans clustering, hierarchical clustering and sentiment analysis to find trend
Normalizing data to reduce variance is necessary preparation of data. Normalizing is rescaling they data to a standard deviation of 1. Sentiment analysis can analyze text content for negative or positive words.
May 5, 2020 • 10min
K means clustering
Unsupervised learning
May 2, 2020 • 33min
Hidden order - understanding the reasons for reinforced learning
Consider a many-player game played over and over. Each player keeps changing his strategy until no further change will make him better off. Equilibrium is reached when each player has chosen a strategy that is optimal for him, given the strategies that the other players are following.
Apr 30, 2020 • 22min
How to build actionable machine learning decisions in production
Everything starts with inference about a business situation . It then moves to observation and experimentation. An finally if an case for impact can be built into ml production either through automation or detection
Apr 28, 2020 • 24min
Logistic regression vs support vector machine
C controls the degree of regularization. Gamma controls the smoothness of the boundary. Kernel can improve speed. Penalty controls the loss function
Apr 25, 2020 • 24min
Percent changed time series and panda groupby and enumerate
I talk about setting up your regression data to be classified using np arrays and dataframes
Apr 22, 2020 • 19min
Gradient boost and adboostingclassifier and bagging and random forest classifier
Improving the decision tree
Apr 22, 2020 • 10min
Variance and bias reflect overfit and underfit in the network
Finding function and ignoring noise to get accurate predictions for unseen data
Apr 20, 2020 • 7min
Linear regressor tree classifier
Information gain and less entropy
Apr 19, 2020 • 26min
Analysis of the profit zone and factors for measuring growth and asset value by sales revenue
1. Who are my customers?
2. How are their priorities changing?
3. Who should be my customer?
4. How can I add value to the customer?
5. How can I become the customers first choice?
6. What is my profit model?
7. What is my current business design?
8. Who are my real competitors?
9. What is my competitors business design?
10 What is my next business design?
11. What is my strategic control point?
12. What is my company worth?
