Stan Stanpol

Competitive Machine Learning

feature engineering (most important by far)!!!!!
simple models
overfitting leaderboard
ensembling

predict the right thing!
build pipeline and put something on the leaderboard
allocate time to play with data, explore
make heavy use of forums
understand subtleties of algos, know what tool to use when

This shows the movement of 330 vehicles travelling round the UK & Ireland over a 24hr period.

Someone asked how to get the latlong from a specific road near a town on OpenStreetMap.

If you need to do it only once (e.g., you're about to go on a trip, and your GPS cannot find your destination city, but allows you to enter GPS coordinates), you can use Nominatim, OpenStreetMap's geocoding interface.

If you need to do it multiple times, in a programmatic manner, there are at least two ways to do that.

Note: I worked with OSM data a couple of years ago, but I don't have an OSM database on my local laptop right now, so some instructions will be a bit fuzzy. I do apologize in advance.

PostGIS queries on a local OSM DB

	"""
	Here is a dice loss for keras which is smoothed to approximate a linear (L1) loss.
	It ranges from 1 to 0 (no error), and returns results similar to binary crossentropy
	"""

	# define custom loss and metric functions

	from keras import backend as K

	def dice_coef(y_true, y_pred, smooth=1):

	'''This script goes along the blog post
	"Building powerful image classification models using very little data"
	from blog.keras.io.
	It uses data that can be downloaded at:
	https://www.kaggle.com/c/dogs-vs-cats/data
	In our setup, we:
	- created a data/ folder
	- created train/ and validation/ subfolders inside data/
	- created cats/ and dogs/ subfolders inside train/ and validation/
	- put the cat pictures index 0-999 in data/train/cats

	{
	"AWSEBDockerrunVersion": "1",
	"Image": {
	"Name": "<AWS_ACCOUNT_ID>.dkr.ecr.us-east-1.amazonaws.com/<NAME>:<TAG>",
	"Update": "true"
	},
	"Ports": [
	{
	"ContainerPort": "443"
	}

	# ### Interface cuSOLVER PyCUDA

	from __future__ import print_function
	import pycuda.gpuarray as gpuarray
	import pycuda.driver as cuda
	import pycuda.autoinit
	import numpy as np
	import scipy.sparse as sp
	import ctypes

	<!DOCTYPE html>
	<html>
	<head>
	<title>Leaflet</title>
	<link rel="stylesheet" href="http://cdn.leafletjs.com/leaflet-0.3.1/leaflet.css" />
	<script src="http://cdn.leafletjs.com/leaflet-0.3.1/leaflet.js"></script>
	<script src="http://maps.google.com/maps/api/js?v=3.2&sensor=false"></script>
	<script src="http://matchingnotes.com/javascripts/leaflet-google.js"></script>
	</head>
	<body>

	#!/usr/bin/env python

	__version__ = "0.00001"

	import sys
	import optparse

	try:
	import Foundation
	import WebKit