Lecture 9 Reinforcement learning
Page 34, loss function, missing the closing bracket (y_i - Q()^2)
Page 21, the second equation of Newton's law is wrong, the physical dimension mismatches on two sides of the equality.
| add_cus_dep('glo', 'gls', 0, 'makeglo2gls'); | |
| add_cus_dep('acn', 'acr', 0, 'makeglo2gls'); | |
| add_cus_dep('ntn', 'not', 0, 'makeglo2gls'); | |
| add_cus_dep('nlo', 'nls', 0, 'make_nomencl'); | |
| sub make_nomencl { | |
| my ($base_name, $path) = fileparse( $_[0] ); | |
| pushd $path; | |
| system "makeindex", "$base_name.nlo", "-s", "nomencl.ist", "-o", "$base_name.nls"; | |
| popd; | |
| } |
| ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQCdypkFlCzvO0eqS7KWUAtPv3XI1ZcSRQNWbMWyMy8KkAxXEkRcULJmvhDNMU0fuKyOFQ/KRFqokCH3zO1HGKFY4gGGq7SnfoBNUVEa9iwtyiKI3MZ9B652T3oOi43QnbmJ/clkk | |
| u6700Ffb5mqmjx5gYSH5LOKRp6ieNz1Qv8m0eflHbR53R8QG5h+xaWk/Va9K8LEMEIsbj5/qlsuOuM/8xzFfDZODdHv9dfo2T6/gQTOPMwK3SG9XAi0sA1uIHmSBV6BpRs1dc1fueyyn5cD7NMD+yagid+lp2CCQhLjH7Cl8hNiBD | |
| t079lco2uFstVN+HarPPGMAt+/PIzCJu3R paris | |
| ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQDRImvL7tqAWHJU2CDRNc/USte9wC+EftvgryrHKpVA0auOBQxUYntkEQugu/2ZWRSzoE9L8pkwJI/O6grYq3mPsRKhQH5C5nFYVM756GNm1Z+D+/GzxruLPCqWnchv42puphosMFtbTN1PEV/pIzKlpFJ4S6E6wKErWI7KYRKqVyAm+ps5UlIsRL9kB6g08U1QBdoG8LRB6iRSVv3nLWdldm1trVt9fNQ1+eSH143wiXVPynUttuSsl8hIfsO/L7ROd6Q6KHSVKvK9lVY+/yp0TICBkUMbM82xmsrLOR7LcxO3dXuLUDBpWIirC/MLS5/PgU1LxwEhBivlAKFCwluN |
| #include <random> | |
| #include <iostream> | |
| int main() { | |
| std::random_device dev; | |
| std::default_random_engine generator(dev()); | |
| std::uniform_real_distribution<double> distribution(0, 1); | |
| for (int i = 0; i < 10; i++) { | |
| double random_number = distribution(generator); | |
| std::cout << random_number << std::endl; | |
| } |
| #!/bin/bash | |
| if [ -f "$1" ]; then | |
| scp ./$1 feng@10.8.6.22:/home/feng/Downloads/ | |
| else | |
| scp feng@10.8.6.22:/home/feng/Downloads/$1 ./ | |
| fi |
| --AppleScript direct | |
| -- set val to do shell script "sed 's|[a-zA-Z]*.pdf|build/&|g' <<< " & quoted form of #PDFPATH# | |
| -- tell application "Skim" to open val | |
| set val to do shell script "sed 's|[a-zA-Z]*.pdf|build/&|g' <<< " & quoted form of #PDFPATH# | |
| -- val is pdfpath | |
| set lf to linefeed | |
| tell application "TeXShop" | |
| set offs to offset of the selection of document #DOCUMENTNAME# |
| import time | |
| import subprocess | |
| from subprocess import PIPE, STDOUT | |
| import sys | |
| from PIL import Image | |
| resolution = 144 | |
| pdf_file_name = sys.argv[1] | |
| r = subprocess.run(['gs', '-q', '-dBATCH', '-dNOPAUSE', '-sDEVICE=bbox', pdf_file_name], stdout=PIPE, stderr=STDOUT) |
Lecture 9 Reinforcement learning
Page 34, loss function, missing the closing bracket (y_i - Q()^2)
Page 21, the second equation of Newton's law is wrong, the physical dimension mismatches on two sides of the equality.
| Page 93: the position of ".text:00000009 lea eax, [ecx+eax*2]" is wrong | |
| Page 59: 0x0000BEEF should be red with the corresponding label |
ITP of prophet: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=960471
叶老师,第32页第13行您提到的exp(2 * beta(B_i - A_i))条件期望和非条件期望同量阶。我尝试仿照(65)证了一下没证出来,可否说明一下证明思路。
| import os | |
| import argparse | |
| def convert(filename): | |
| with open(filename, 'rb') as f: | |
| st = f.read() | |
| try: | |
| st = st.decode('gbk') | |
| except Exception as e: | |
| print('gbk decode error for ' + filename) |