mirror of
https://github.com/Andreaierardi/Master-DataScience-Notes.git
synced 2024-11-11 14:29:31 +01:00
43 lines
2.7 KiB
TeX
43 lines
2.7 KiB
TeX
\babel@toc {english}{}
|
||
\contentsline {chapter}{\numberline {1}Lecture 1 - 09-03-2020}{4}%
|
||
\contentsline {section}{\numberline {1.1}Introduction}{4}%
|
||
\contentsline {chapter}{\numberline {2}Lecture 2 - 07-04-2020}{7}%
|
||
\contentsline {section}{\numberline {2.1}Argomento}{7}%
|
||
\contentsline {section}{\numberline {2.2}Loss}{7}%
|
||
\contentsline {subsection}{\numberline {2.2.1}Absolute Loss}{7}%
|
||
\contentsline {subsection}{\numberline {2.2.2}Square Loss}{8}%
|
||
\contentsline {subsection}{\numberline {2.2.3}Example of information of square loss}{8}%
|
||
\contentsline {subsection}{\numberline {2.2.4}labels and losses}{9}%
|
||
\contentsline {subsection}{\numberline {2.2.5}Example TF(idf) documents encoding}{11}%
|
||
\contentsline {chapter}{\numberline {3}Lecture 3 - 07-04-2020}{13}%
|
||
\contentsline {section}{\numberline {3.1}Overfitting}{15}%
|
||
\contentsline {subsection}{\numberline {3.1.1}Noise in the data}{15}%
|
||
\contentsline {section}{\numberline {3.2}Underfitting}{16}%
|
||
\contentsline {section}{\numberline {3.3}Nearest neighbour}{17}%
|
||
\contentsline {chapter}{\numberline {4}Lecture 4 - 07-04-2020}{19}%
|
||
\contentsline {section}{\numberline {4.1}Computing $h_{NN}$}{19}%
|
||
\contentsline {section}{\numberline {4.2}Tree Predictor}{20}%
|
||
\contentsline {chapter}{\numberline {5}Lecture 5 - 07-04-2020}{23}%
|
||
\contentsline {section}{\numberline {5.1}Tree Classifier}{23}%
|
||
\contentsline {section}{\numberline {5.2}Jensen’s inequality}{24}%
|
||
\contentsline {section}{\numberline {5.3}Tree Predictor}{26}%
|
||
\contentsline {section}{\numberline {5.4}Statistical model for Machine Learning}{27}%
|
||
\contentsline {chapter}{\numberline {6}Lecture 6 - 07-04-2020}{29}%
|
||
\contentsline {section}{\numberline {6.1}Bayes Optimal Predictor}{29}%
|
||
\contentsline {subsection}{\numberline {6.1.1}Square Loss}{30}%
|
||
\contentsline {subsection}{\numberline {6.1.2}Zero-one loss for binary classification}{31}%
|
||
\contentsline {section}{\numberline {6.2}Bayes Risk}{33}%
|
||
\contentsline {chapter}{\numberline {7}Lecture 7 - 07-04-2020}{34}%
|
||
\contentsline {section}{\numberline {7.1}Chernoff-Hoffding bound}{34}%
|
||
\contentsline {section}{\numberline {7.2}Union Bound}{35}%
|
||
\contentsline {section}{\numberline {7.3}Studying overfitting of a ERM}{39}%
|
||
\contentsline {chapter}{\numberline {8}Lecture 8 - 07-04-2020}{41}%
|
||
\contentsline {section}{\numberline {8.1}The problem of estimating risk in practise}{42}%
|
||
\contentsline {section}{\numberline {8.2}Cross-validation}{44}%
|
||
\contentsline {section}{\numberline {8.3}Nested cross validation}{46}%
|
||
\contentsline {chapter}{\numberline {9}Lecture 9 - 07-04-2020}{47}%
|
||
\contentsline {section}{\numberline {9.1}Tree predictors}{47}%
|
||
\contentsline {subsection}{\numberline {9.1.1}Catalan Number}{49}%
|
||
\contentsline {chapter}{\numberline {10}Lecture 10 - 07-04-2020}{53}%
|
||
\contentsline {section}{\numberline {10.1}TO BE DEFINE}{53}%
|