rreas

function [U,V,idx] = spkmeans(X,k,tol,imax)

  [d,n] = size(X);
  
  U = zeros(d,k);
  V = zeros(k,n);
  
  % random clusters and normalize to unit sphere.
  for j = 1:n
    V(randi(k),j) = 1;

KWMalik

##Google Interview Questions: Product Marketing Manager

• Why do you want to join Google? -- Because I want to create tools for others to learn, for free. I didn't have a lot of money when growing up so I didn't get access to the same books, computers and resources that others had which caused money, I want to help ensure that others can learn on the same playing field regardless of their families wealth status or location.
• What do you know about Google’s product and technology? -- A lot actually, I am a beta tester for numerous products, I use most of the Google tools such as: Search, Gmaill, Drive, Reader, Calendar, G+, YouTube, Web Master Tools, Keyword tools, Analytics etc.
• If you are Product Manager for Google’s Adwords, how do you plan to market this?
• What would you say during an AdWords or AdSense product seminar?
• Who are Google’s competitors, and how does Google compete with them? -- Google competes on numerous fields: --- Search: Baidu, Bing, Duck Duck Go

tsiege

ANNOUNCEMENT

I have moved this over to the Tech Interview Cheat Sheet Repo and has been expanded and even has code challenges you can run and practice against!






\

kevin-keraudren

#!/usr/bin/python

import SimpleITK as sitk
import vtk
import numpy as np
import sys

from vtk.util.vtkConstants import *

filename = sys.argv[1]

debasishg

1. Feature Learning

• Learning Feature Representations with K-means by Adam Coates and Andrew Y. Ng
• The devil is in the details: an evaluation of recent feature encoding methods by Chatfield et. al.
• Emergence of Object-Selective Features in Unsupervised Feature Learning by Coates, Ng
• Scaling Learning Algorithms towards AI Benjio & LeCun
• A Theory of Feature Learning by Brendan van Rooyen, Robert C. Williamson

2. Deep Learning

• Dropout: A Simple Way to Prevent Neural Networks from Overfitting by Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever and Ruslan Salakhutdinov
• [Understanding

karpathy

"""
This is a batched LSTM forward and backward pass
"""
import numpy as np
import code

class LSTM:
  
  @staticmethod
  def init(input_size, hidden_size, fancy_forget_bias_init = 3):

myungsub

Vision & Language

• Ask Your Neurons: A Neural-Based Approach to Answering Questions About Images

  • Mateusz Malinowski, Marcus Rohrbach, Mario Fritz

• Aligning Books and Movies: Towards Story-Like Visual Explanations by Watching Movies and Reading Books

  • Yukun Zhu, Ryan Kiros, Rich Zemel, Ruslan Salakhutdinov, Raquel Urtasun, Antonio Torralba, Sanja Fidler

• Learning Query and Image Similarities With Ranking Canonical Correlation Analysis

• Wah Ngo

saliksyed

""" Deep Auto-Encoder implementation
	
	An auto-encoder works as follows:

	Data of dimension k is reduced to a lower dimension j using a matrix multiplication:
	softmax(W*x + b)  = x'
	
	where W is matrix from R^k --> R^j

	A reconstruction matrix W' maps back from R^j --> R^k

atduskgreg

Interactive Machine Learning

Taught by Brad Knox at the MIT Media Lab in 2014. Course website. Lecture and visiting speaker notes.

• Power to the People: The Role of Humans in Interactive Machine Learning by Knox, Cakmak, Kulesza, Amershi, and Lau
• A Few Useful Things to Know about Machine Learning by Domingos
• Machine Learning that Matters by Wagstaff
• Beyond Concise and Colorful: Learning Intelligible Rules by Pazzani et al.
• [Designing Games with a Purpose] (https://www.cs.cmu.edu/~biglou/GWAP_CACM.pdf) by von Ahn and Dabbish
• [Human Model Evaluation in Interactive Supervised

kashif

Cross Entropy Method

How do we solve for the policy optimization problem which is to maximize the total reward given some parametrized policy?

Discounted future reward

To begin with, for an episode the total reward is the sum of all the rewards. If our environment is stochastic, we can never be sure if we will get the same rewards the next time we perform the same actions. Thus the more we go into the future the more the total future reward may diverge. So for that reason it is common to use the discounted future reward where the parameter discount is called the discount factor and is between 0 and 1.

A good strategy for an agent would be to always choose an action that maximizes the (discounted) future reward. In other words we want to maximize the expected reward per episode.