Mercurial > hg > hybrid-music-recommender-using-content-based-and-social-information
view poster/sections/method2.tex @ 47:b0186d4a4496 tip
Move 7Digital dataset to Downloads
author | Paulo Chiliguano <p.e.chiliguano@se14.qmul.ac.uk> |
---|---|
date | Sat, 09 Jul 2022 00:50:43 -0500 |
parents | fcd3730ffc19 |
children |
line wrap: on
line source
\section{Data Representation} \begin{itemize} \item The CNN produces a vector representation of the spectrograms for each audio file. \item The EDA \cite{Ding2015451} estimates a probabilistic model from a sample to model the user profiles from the user ratings dataset and vector representation. \end{itemize} %$$ %\label{eq:pgmmx} %p(x)=\sum^{K}_{k=1}p(x|k)P(k) %$$ \begin{center} \resizebox*{0.85\columnwidth}{!}{\includegraphics{images/diagram_cnn.eps}}\\ {\large \textbf{Fig. 2.} Architecture of a CNN for music genre representation}\\[2ex] \resizebox*{0.6\columnwidth}{!}{\includegraphics{images/flowchart_eda.eps}}\\ {\large \textbf{Fig. 3.} General flowchart for an EDA}\\ %\begin{tabular}{c@{ }c} %\resizebox*{0.45\columnwidth}{!}{\includegraphics{images/diagram_hybrid_music_recommender.eps}} & %\resizebox*{0.45\columnwidth}{!}{\includegraphics{images/fruit/plums.eps}} \\ %Bananas & Plums \\ %\end{tabular} \end{center}