2018-12 在职刷题保持状态， 每天1题+ML insights

Day39

Had to spend most of time writing my paper.

Solve three ez problems.

Solved 203, 973 and 339.

Day40

Solved 764 with DP.
Solved 750 with set. Surprisingly, naive solution also passed...
Solved 745 with HashTable


Planned paper revision.

Day41

Solved 714 with DP.
Solved 721 with Union-find.

Solved 674.

Day42

Solved 946.
Solved 744.

Solved 366.

Day43

Solved 698 with DFS. Surprised that the DFS solution passed. It is at least an exponential solution though.
Solved 633, 671.

Submit my paper to Nature Communications. Finger crossed.

Day44.

Got interview chance from Linkedin ML team. Put some efforts to review my ML skillset.

Solved 244 with Binary Search

Solved 364.

Solved 605.
Reviewed:

-Generalized Linear Models and exponential family
-Gaussian Discriminant Analysis, Naive Bayes.
-Regression/Classification Trees. Impurity Measures (Entropy, Misclassification, Gini Index)
-Boot Strapping aggregation/Bagging.

Day45

ML is much more fun than solving LC problems...

Solved 170.
Solved 528. Classic russian roulette sampling of a weighted distribution.

Solved 432 with doubly linked list and hash table. The longest solution ever written for LC. Many boundary conditions to consider.
The key point is to have an organized thinking about inc and dec.  First update the new node to which the key is shifted. The update the old node where the key was in.

Day46

Solved 464 with memoization+minmax recursion+binary state

Solved 367

Solved 515

Day47

Solved 237.
Solved 272.
Solved 254.

Day48
ML reviewed:
Impurity measure + decision theory
SVM, Kernel method, Lagrange multiplier.

Solved 156.
Solved 277.
Solved 311