ajtulloch

predict' :: DecisionTree -> V.Vector Double -> Double
predict' (Leaf value) _ = value
predict' (Branch feature splitValue left right) featureVector =
    if featureVector ! feature < splitValue
    then predict' left featureVector
    else predict' right featureVector

ajtulloch

=================================================================================================
                    Compiled              Naive         Piecewise Flattened  Contiguous Flattened
-------------------------------------------------------------------------------------------------
(Intercept)      -11560.208***        -8697.547***         -11962.803***        -18324.924***    
                  (1436.562)          (1105.108)            (2250.144)           (3568.027)      
num_trees            54.382***           44.813***             44.427***            64.475***    
                     (6.300)             (4.846)               (9.868)             (15.647)      
depth              2479.344***         1975.124***           2441.172***          3651.876***    
                   (233.106)           (179.322)             (365.123)            (578.972)      
num_features          0.058***            0.074***              0.166***             0.268***    

ajtulloch

library("nlme")
library("data.table")
library("memisc")
library("functional")

kColumns <- c("algorithm", "num_trees", "depth", "num_features", "time_ns")

preprocess <- function(filename) {
  # Could use read.table, but fails for some reason
  df <- data.table(read.csv(filename, col.names=kColumns))

ajtulloch

\documentclass[12pt]{amsart}
\usepackage{amsthm, amsmath, amssymb}
\usepackage{setspace}
\usepackage{listings}
\onehalfspacing                 

\theoremstyle{plain}% default 
\newtheorem{thm}{Theorem}[section] 
\newtheorem{lem}[thm]{Lemma} 
\newtheorem{prop}[thm]{Proposition} 

ajtulloch

library('ProjectTemplate')
load.project()

## Linear Regression
mod <- lm(Y ~ ., data = zip.train.filtered)

# Round predictions
category_f <- function(x) { if (x > 2.5) 3 else 2 }
predictions.lm.test <- as.character(sapply(predict(mod, zip.test.filtered), 
                                           category_f))

ajtulloch

Simple Examples

This section introduces the usage of the LaTeX2Markdown tool, showing an example of the various environments available.

Theorem 1 (Euclid, 300 BC)

There are infinitely many primes.

ajtulloch

GHCi, version 7.6.3: http://www.haskell.org/ghc/  :? for help
Loading package ghc-prim ... linking ... done.
Loading package integer-gmp ... linking ... done.
Loading package base ... linking ... done.
Prelude> :cd /Users/tulloch/Code/haskell/machinelearning/
Prelude> :load "MachineLearning/HopfieldDemonstration.hs"
[2 of 3] Compiling MachineLearning.Hopfield ( MachineLearning/Hopfield.hs, interpreted )
[3 of 3] Compiling MachineLearning.HopfieldDemonstration ( MachineLearning/HopfieldDemonstration.hs, interpreted )
Ok, modules loaded: MachineLearning.HopfieldDemonstration, MachineLearning.Hopfield, MachineLearning.Util.
*MachineLearning.HopfieldDemonstration> main

ajtulloch

import numpy as np
from collections import namedtuple

import logging
logging.basicConfig(level=logging.DEBUG)


class BarrierFunction(object):
    def __init__(self, n):
        pass

ajtulloch

import numpy as np
from collections import namedtuple

import logging
logging.basicConfig(level=logging.DEBUG)


class BarrierFunction(object):
    def __init__(self, n):
        pass

ajtulloch

import numpy as np
import cvxopt


class Kernel(object):
    """Implements list of kernels from
    http://en.wikipedia.org/wiki/Support_vector_machine
    """
    @staticmethod
    def linear(x, y):