Jason Liu jxnl

About myself, and work history

Founded Data Hackers, Data Science Club w/ Kevin Huang and Anni Dong.

Sysomos
Nyu
Stitchfix
Facebook

Disclaimer

Mentors
Conceptual Heavyweights
Trends visionaries
New perspectives
Client point person
Key player in business development
Questions everything - Designs Solutions Goals
Push/Review Designers
Risk Taking
Moral

2cents on diversity.

Some friends have asked me about my thoughts on google memo.

Stereotype threat is an observable phenomena.

It doesn’t explain every performance gap, and eliminating it won’t close all performance gaps, however, it’s net loss and suboptimal.

tl;dr

motivation

Variational Autoencoders (VAE) are extremely appealing as they allow for learning complicated distributions taking advantage of recent progress in gradient descent algorithms and accelerated processing with GPUs. The latent space of regular autoencoders are typically very sparse and unrestricted, making it difficult to generate data robust to variations of the latent variables. Traditional VAEs use normal priors regularize the latent space however these assuptions are unsound when it comes to complex data such as images or text.

challenge

We propose Dirichlet priors for a multinomial latent space. This latent space allows us to explore the data by interpreting $P(\theta_i)$ as the likelihood of belonging to a latent class, as a form of non-linear soft clustering.

	t = """
	paste transcript
	"""


	grades = [x for x in t.split("\n\n") if "Y" in x][1:]
	grades =[[x for x in y.strip().replace(" ", "_").replace(" ", "_").split("_") if x != ""][:4] for y in grades]
	grades = [(x[0], x[-1]) for x in grades if len(x[-1]) == 2]
	df = pd.DataFrame(grades, columns=["course_code", "grade"])

	log - loading features
	feature: train
	feature: p.h.building
	feature: p.m.building
	feature: p.l.building
	feature: p.h.manager
	feature: p.m.manager
	feature: p.l.manager
	feature: n_photos
	feature: latitude


	# coding: utf-8

	# # Loading and Transforming json data

	# In[1]:

	import pandas as pd
	import numpy as np
	import pickle

	import mechanize
	import pandas as pd
	import os

	from bs4 import BeautifulSoup
	from IPython.core.display import display, HTML

	br = mechanize.Browser()

	br.open("http://www.adm.uwaterloo.ca/infocour/CIR/SA/under.html")

	def simpson(a, b, f, N):
	return (1.0 / 3.0) * (2 * (((b - a) / N) * sum(f(v) for v in [i * ((b - a) \
	/ (2. * N)) for i in range(2 * N + 1)][1:2 * N + 1:2])) + (((b - a) / N) * \
	((f(a) + f(b)) / 2.0 + sum (f(v * ((b - a) / N) + a) for v in xrange(1, N)))
	))