This is an OpenPGP proof that connects my OpenPGP key to this Github account. For details check out https://keyoxide.org/guides/openpgp-proofs
[Verifying my OpenPGP key: openpgp4fpr:82E707B0D7283273CB9D53969596671D5E50863D]
This is an OpenPGP proof that connects my OpenPGP key to this Github account. For details check out https://keyoxide.org/guides/openpgp-proofs
[Verifying my OpenPGP key: openpgp4fpr:82E707B0D7283273CB9D53969596671D5E50863D]
\documentclass{minimal} % Default font size and paper size | |
\usepackage{fontspec} % For loading fonts | |
\setmainfont[RawFeature={-calt}, Renderer=Harfbuzz]{PragmataPro Liga}[ | |
UprightFont={*Regular}, | |
ItalicFont={*Italic}, | |
BoldFont={*Bold}, | |
BoldItalicFont={*Bold Italic}, | |
] |
using DiffEqFlux | |
using Zygote | |
nn = FastChain((x,p) -> p) | |
p = rand(2, 2) | |
x = rand(1, 100) | |
function f(p) | |
gz, back = Zygote.pullback(z -> nn(z, p), x) | |
back(gz)[1] |
julia> CuArrays.zeros(128, 32) |> nn | |
┌ Warning: calls to Base intrinsics might be GPU incompatible | |
│ exception = | |
│ You called exp(x::T) where T<:Union{Float32, Float64} in Base.Math at special/exp.jl:75, maybe you intended to call exp(x::Float32) in CUDAnative at /home/guillaume/.julia/packages/CUDAnative/hfulr/src/device/cuda/math.jl:101 instead? | |
│ Stacktrace: | |
│ [1] exp at special/exp.jl:75 | |
│ [2] mish at /home/guillaume/.julia/packages/NNlib/FAI3o/src/activation.jl:206 | |
│ [3] #25 at /home/guillaume/.julia/packages/GPUArrays/1wgPO/src/broadcast.jl:49 | |
└ @ CUDAnative ~/.julia/packages/CUDAnative/hfulr/src/compiler/irgen.jl:111 | |
┌ Warning: calls to Base intrinsics might be GPU incompatible |
# -*- coding: utf-8 -*- | |
import scrapy | |
import datetime | |
import json | |
from TwitterScraper.items import Tweet | |
from urllib.parse import quote | |
from bs4 import BeautifulSoup | |
from scrapy.http import HtmlResponse | |
from dateutil.parser import parse |
#!/usr/bin/env python3 | |
import collections | |
import copy | |
import datetime | |
import functools | |
import itertools | |
import json | |
import lzma | |
import gzip |
mxnet ) R CMD INSTALL mxnet_0.5.tar.gz | |
* installing to library ‘/home/guillaume/R/x86_64-pc-linux-gnu-library/3.2’ | |
* installing *source* package ‘mxnet’ ... | |
** libs | |
g++-4.9 -I/usr/share/R/include -DNDEBUG -I../inst/include -I"/home/guillaume/R/x86_64-pc-linux-gnu-library/3.2/Rcpp/include" -fpic -g -O2 -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -g -c executor.cc -o executor.o | |
g++-4.9 -I/usr/share/R/include -DNDEBUG -I../inst/include -I"/home/guillaume/R/x86_64-pc-linux-gnu-library/3.2/Rcpp/include" -fpic -g -O2 -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -g -c export.cc -o export.o | |
g++-4.9 -I/usr/share/R/include -DNDEBUG -I../inst/include -I"/home/guillaume/R/x86_64-pc-linux-gnu-library/3.2/Rcpp/include" -fpic -g -O2 -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -g -c io.cc -o io.o | |
g++-4.9 -I/usr/share/R/include -DNDEBUG -I../inst/include -I"/home/guillaume/R/x86_64-pc-linux-gnu-library/3.2/Rcpp |
\section{Théorie} | |
\subsection{L'apprentissage machine} | |
Initialement une branche des statistiques, l'apprentissage statistique s'est rapidement transformé en une discipline à part entière mêlant plusieurs domaines des mathématiques et de l'informatique: l'apprentissage machine. | |
Le terme \emph{apprentissage statistique} en lui-même est vague et regroupe plusieurs sous-domaines. De façon générale on dispose d'un échantillon $\mathcal{L}$ d'individus possédant des caractéristiques $X_i \in \mathcal{X}$ propres considérées comme déterministes appelées variables et un attribut aléatoire $Y \in \mathcal{Y}$. Si $\mathcal{Y}$ est un ensemble discret on parle de problème de \emph{classification}, s’il est continu on parle alors de problème de \emph{régression}. Il existe un grand nombre d'autres objectifs comme le \emph{clustering}, la \emph{détection de structures} et autres, mais nous ne nous intéresserons ici qu'à ces deux grandes familles en choisissant à chaque fois la tache qui facilite les explications ou es |
double bsc(double x, double T, double K,double L, double r,double sigma) { | |
double lambda = (r+sigma*sigma*0.5)/(sigma*sigma) ; | |
double x1 = log(x/L)/(sigma*sqrt(T)) + lambda * sigma * sqrt(T) ; | |
double y1 = log(L/x)/(sigma*sqrt(T)) + lambda * sigma * sqrt(T) ; | |
double d1 = ( log(x/K) + (r+sigma*sigma*0.5)*T ) / (sigma * sqrt(T)) ; | |
double d2 = d1 - sigma * sqrt(T) ; | |
double y = log(L*L/(x*K)) / (sigma * sqrt(T)) + lambda * sigma * sqrt(T) ; | |
double cui = x * N(x1) - K * exp(-r*T) * N(x1 - sigma * sqrt(T)) - x * pow(L/x,2*lambda) * ( N(-y) - N(-y1) ) + K * exp(-r*T) * pow(L/x,2*lambda-2) * ( N(-y+sigma*sqrt(T)) - N(-y1 + sigma*sqrt(T)) ) ; | |
double c = x * N(d1) - K* exp(-r*T) * N(d2) ; | |
return(c - cui ); |
*{-webkit-box-sizing:border-box;-moz-box-sizing:border-box;box-sizing:border-box;color:#839496}::-moz-selection{background:#FF5E99;color:#fff;text-shadow:none}::selection{background:#FF5E99;color:#fff;text-shadow:none}a, a:active, a:visited{color:#b58900}a:link{-webkit-tap-highlight-color:#FF5E99}body{line-height:1.5;font-size:22px;background:#fdf6e3;color:#839496 font-family:"Gentium",Georgia,"Times New Roman",Times,serif}strong{font-weight:bold}dt{font-weight:bold}h1{font-size:1.6em}h1 a:hover{-webkit-text-stroke:3px #073642}h2{font-size:1.3em}h3{font-size:1em}h1, h2, h3, h4, h5, h6{color:#657b83;border-color:#839496;font-weight:normal;font-family:"Open Sans", sans-serif}td{padding:8px}blockquote{font-style:italic;background:#eee8d5;padding:5px 20px;border-radius:5px;-moz-border-radius:5px;-webkit-border-radius:5px}.title{text-align:center}.large{font-size:5em}.container{margin:10px;padding-bottom:50px;color:#657b83}.header{margin:auto;font-size:16px;display:block;text-align:left}.content{width:60%;margin:0 |