Main Page and Deep Learning: Difference between pages

From EMC23 - Satellite Of Love
(Difference between pages)
Jump to navigation Jump to search
 
mNo edit summary
 
Line 1: Line 1:
<strong>Electronic Music Coders Amsterdam</strong>
Deep Learning is a subset of [[Machine Learning]] (which would also include Reinforcement learning)


https://www.meetup.com/Electronic-Music-Coding
= Applications =
* Speech Recognition
* Voice Based emotion classification
* Noise recognition
* Musical Genre Instrument Mood Classificatiob
* Music Tagging
* Music Generation


== The Group ==
= Definitions =
Deep learning is a class of machine learning algorithms that[12](pp199–200) uses multiple layers to progressively extract higher-level features from the raw input.


* [https://www.meetup.com/Electronic-Music-Coding/ Electronic Music Coding Amsterdam]
Learning can be supervised, semi-supervised or unsupervised


* [http://www.emc23.com The Blog]
'''FFNN''': The most straightforward way of structuring a neural network, the Feed Forward Neural Network (FFNN) structures neurons in a series of layers, with each neuron in a layer containing weights to all neurons in the previous layer.<br />
* [http://wiki.emc23.com This Wiki]
'''CNN''': Convolutional Neural Networks (CNN) represent nested filters over grid-organized data. They are by far the most commonly used type of model when processing images.<br />
* [https://twitter.com/emc23dotcom/ Twitter Account]
'''RNN/LSTM''': Recurrent Neural Networks (RNN) and the related Long Short-Term Memory (LSTM) model types are structured to effectively represent for loops in traditional computing, collecting state while iterating over some object. They can be used for processing sequences of data.<br />
* [https://github.com/EMC23 Github Account]
'''Transformer''': A more modern replacement for RNN/LSTMs, the transformer architecture enables training over larger datasets involving sequences of data.<br />


== The Topics ==
= Python =
[[Coding]], [[Live Coding]], [[Studio Technology]], [[Installations]], [[Laptop Jamming]], [[Modular]], [[DIY]], [[Composition]], [[Digital Instruments]], [[Creative Coding]], [[Generative Music]], [[Deep Learning]], [[DSP]], [[Graphic Interfaces]], [[DJing]] and [[Game Music]]
[[Librosa]] is used to analyse and manipulate audio


== The Goal==
[[Tensorflow]] is used to train models
Meet, Share Knowledge, Network.


[[Keras]] High Level Library for Tensorflow


== The Projects==
Accuracy    False positive rate Precision


=== Daily - Audio/Visual Streaming ===
Confusion matrix False negative rate Recall
[[OBS]] and [[MSDP]] and Ableton [[Max4Live]] Electronic Audio/visual stream


This is where we connect Ableton Live to a [[Node.js]] server via [[Max]] , Midi sync it to MSDP and pipe it to OBS  which combines the video with [[VLC]] video footage. The video sources are as diverse as the Archive.org, UBU, open culture and social media platforms.
F1 ScoreLog LossROC curve


Always looking for new show material and ideas.
Negative predictive value Specificity


=== Weekly - Friday Jamming Session ===
= Deep learning  architectures used for music generation =


Ableton Link and/or midi sync up. Every genre investigated. Participants take turns in groups of varying size
From this basic building block, we will describe in the following sections the main types of deep learning architectures used for music generation (as well as for other purposes):


=== Evening Study Classes ===
• feedforward,


Tues, Wed, Thurs
• autoencoder,
1 on 1 Audio Programming Workshops
* [[Javascript]] ([[Web Audio]])
* [[Pure Data]] or [[Max]]/Msp (students choce)
* [[C++]] (Juice or VULT) [[VCVRack]] or [[VST]] audio plugin
* [[Python]] Machine Learning
* Hardware [[Bela]] or [[Arduino]] (Students choice)
* [[Livestreaming]]


=== Monthly - Open Session ===
• restricted Boltzmann machine (RBM),
Bandcamp Friday Livestream


=== Workshops ===
• recurrent (RNN).


Connect a midi controller to the browser
= Architectural Patterns =
which could be applied to them)


https://gizmodo.com/a-beginners-guide-to-the-synth-1736978695
• convolutional,


* [[Build a synth in C++]]
• conditioning,
* [[Build a synth in Chrome Browser]] Javascript
* [[Build a synth in VCVRack]]
* [[Build a synth in Reaktor]]
* [[Build a synth in Reaktor Blocks]]
* [[Build a synth in Pure Data]]
* [[Build a synth in Max For Live]]


==Minimum Requirements ==
• adversarial.


* Laptop
= Links =
* [[Asio]] compliant soundcard
Audio Handling Basics: Process Audio Files In Command-Line or Python
* [[Midi Controller]] (will be supplied for first few sessions)
* Bring headphones if you have them


==Cost ==
https://benhayes.net/projects/nws/#audio-examples<br />
*  Evening sessions : 300 Euro per Quarter


==The Place ==
Do Androids Dream of Electric Beats?<br />
the Satellite Of Love
https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53<br />
Intro to Audio Analysis: Recognizing Sounds Using Machine Learning<br />
https://magenta.tensorflow.org/music-vae<br />
https://musicalmetacreation.org/mume2018/proceedings/Sturm.pdf<br />
https://ccrma.stanford.edu/~blackrse/algorithm.html<br />
https://magenta.tensorflow.org/music-vae<br />
https://mirg.city.ac.uk/codeapps/the-magnatagatune-dataset<br />
https://docs.microsoft.com/en-us/cognitive-toolkit/<br />
https://scikit-learn.org/stable/<br />
https://pythonrepo.com/repo/nerdyrodent-VQGAN-CLIP-python-deep-learning<br />


Panamalaan 6d
= Terminology =
Impute is a common term referring to different statistical tools which can be used to calculate missing values from your dataset.
Outliers are data points that are significantly different from others in the same sample.


1019 NE Amsterdam
Glossary
 
Bag of words: A technique used to extract features from the text. It counts how many times a word appears in a document (corpus), and then transforms that information into a dataset.
 
A categorical label has a discrete set of possible values, such as "is a cat" and "is not a cat."
 
Clustering. Unsupervised learning task that helps to determine if there are any naturally occurring groupings in the data.
 
CNN: Convolutional Neural Networks (CNN) represent nested filters over grid-organized data. They are by far the most commonly used type of model when processing images.
 
A continuous (regression) label does not have a discrete set of possible values, which means possibly an unlimited number of possibilities.
 
Data vectorization: A process that converts non-numeric data into a numerical format so that it can be used by a machine learning model.
 
Discrete: A term taken from statistics referring to an outcome taking on only a finite number of values (such as days of the week).
 
FFNN: The most straightforward way of structuring a neural network, the Feed Forward Neural Network (FFNN) structures neurons in a series of layers, with each neuron in a layer containing weights to all neurons in the previous layer.
 
Hyperparameters are settings on the model which are not changed during training but can affect how quickly or how reliably the model trains, such as the number of clusters the model should identify.
 
Log loss is used to calculate how uncertain your model is about the predictions it is generating.
 
Hyperplane: A mathematical term for a surface that contains more than two planes.
 
Impute is a common term referring to different statistical tools which can be used to calculate missing values from your dataset.
 
label refers to data that already contains the solution.
 
loss function is used to codify the model’s distance from this goal
 
Machine learning, or ML, is a modern software development technique that enables computers to solve problems by using examples of real-world data.
 
Model accuracy is the fraction of predictions a model gets right. Discrete: A term taken from statistics referring to an outcome taking on only a finite number of values (such as days of the week). Continuous: Floating-point values with an infinite range of possible values. The opposite of categorical or discrete values, which take on a limited number of possible values.
 
Model inference is when the trained model is used to generate predictions.
 
model is an extremely generic program, made specific by the data used to train it.
 
Model parameters are settings or configurations the training algorithm can update to change how the model behaves.
 
Model training algorithms work through an interactive process where the current model iteration is analyzed to determine what changes can be made to get closer to the goal. Those changes are made and the iteration continues until the model is evaluated to meet the goals.
 
Neural networks: a collection of very simple models connected together. These simple models are called neurons. The connections between these models are trainable model parameters called weights.
 
Outliers are data points that are significantly different from others in the same sample.
 
Plane: A mathematical term for a flat surface (like a piece of paper) on which two points can be joined by a straight line.
 
Regression: A common task in supervised machine learning.
 
In reinforcement learning, the algorithm figures out which actions to take in a situation to maximize a reward (in the form of a number) on the way to reaching a specific goal.
 
RNN/LSTM: Recurrent Neural Networks (RNN) and the related Long Short-Term Memory (LSTM) model types are structured to effectively represent for loops in traditional computing, collecting state while iterating over some object. They can be used for processing sequences of data.
 
Silhouette coefficient: A score from -1 to 1 describing the clusters found during modeling. A score near zero indicates overlapping clusters, and scores less than zero indicate data points assigned to incorrect clusters. A
 
Stop words: A list of words removed by natural language processing tools when building your dataset. There is no single universal list of stop words used by all-natural language processing tools.
 
In supervised learning, every training sample from the dataset has a corresponding label or output value associated with it. As a result, the algorithm learns to predict labels or output values.
 
Test dataset: The data withheld from the model during training, which is used to test how well your model will generalize to new data.
 
Training dataset: The data on which the model will be trained. Most of your data will be here.
 
Transformer: A more modern replacement for RNN/LSTMs, the transformer architecture enables training over larger datasets involving sequences of data.
 
In unlabeled data, you don't need to provide the model with any kind of label or solution while the model is being trained.
 
In unsupervised learning, there are no labels for the training data. A machine learning algorithm tries to learn the underlying patterns or distributions that govern the data.
 
 
 
 
 
 
 
 
https://youtu.be/CNNmBtNcccE
 
 
 
Machine learning is synthesizing death metal. It might make your death metal radio DJ nervous – but it could also mean music software works with timbre and time in new ways. That news – plus some comical abuse of neural networks for writing genre-specific lyrics in genres like country – next.
Peter Kirn http://cdm.link/2019/04/now-ai-takes-on-writing-death-metal-country-music-hits-more/
 
 
 
<evlplayer id="player1" w="480" h="360" service="youtube" defaultid="CNNmBtNcccE" />

Revision as of 22:10, 25 August 2021

Deep Learning is a subset of Machine Learning (which would also include Reinforcement learning)

Applications[edit]

  • Speech Recognition
  • Voice Based emotion classification
  • Noise recognition
  • Musical Genre Instrument Mood Classificatiob
  • Music Tagging
  • Music Generation

Definitions[edit]

Deep learning is a class of machine learning algorithms that[12](pp199–200) uses multiple layers to progressively extract higher-level features from the raw input.

Learning can be supervised, semi-supervised or unsupervised

FFNN: The most straightforward way of structuring a neural network, the Feed Forward Neural Network (FFNN) structures neurons in a series of layers, with each neuron in a layer containing weights to all neurons in the previous layer.
CNN: Convolutional Neural Networks (CNN) represent nested filters over grid-organized data. They are by far the most commonly used type of model when processing images.
RNN/LSTM: Recurrent Neural Networks (RNN) and the related Long Short-Term Memory (LSTM) model types are structured to effectively represent for loops in traditional computing, collecting state while iterating over some object. They can be used for processing sequences of data.
Transformer: A more modern replacement for RNN/LSTMs, the transformer architecture enables training over larger datasets involving sequences of data.

Python[edit]

Librosa is used to analyse and manipulate audio

Tensorflow is used to train models

Keras High Level Library for Tensorflow

Accuracy False positive rate Precision

Confusion matrix False negative rate Recall

F1 ScoreLog LossROC curve

Negative predictive value Specificity

Deep learning architectures used for music generation[edit]

From this basic building block, we will describe in the following sections the main types of deep learning architectures used for music generation (as well as for other purposes):

• feedforward,

• autoencoder,

• restricted Boltzmann machine (RBM),

• recurrent (RNN).

Architectural Patterns[edit]

which could be applied to them)

• convolutional,

• conditioning,

• adversarial.

Links[edit]

Audio Handling Basics: Process Audio Files In Command-Line or Python

https://benhayes.net/projects/nws/#audio-examples

Do Androids Dream of Electric Beats?
https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53
Intro to Audio Analysis: Recognizing Sounds Using Machine Learning
https://magenta.tensorflow.org/music-vae
https://musicalmetacreation.org/mume2018/proceedings/Sturm.pdf
https://ccrma.stanford.edu/~blackrse/algorithm.html
https://magenta.tensorflow.org/music-vae
https://mirg.city.ac.uk/codeapps/the-magnatagatune-dataset
https://docs.microsoft.com/en-us/cognitive-toolkit/
https://scikit-learn.org/stable/
https://pythonrepo.com/repo/nerdyrodent-VQGAN-CLIP-python-deep-learning

Terminology[edit]

Impute is a common term referring to different statistical tools which can be used to calculate missing values from your dataset. Outliers are data points that are significantly different from others in the same sample.


Glossary


Bag of words: A technique used to extract features from the text. It counts how many times a word appears in a document (corpus), and then transforms that information into a dataset.

A categorical label has a discrete set of possible values, such as "is a cat" and "is not a cat."

Clustering. Unsupervised learning task that helps to determine if there are any naturally occurring groupings in the data.

CNN: Convolutional Neural Networks (CNN) represent nested filters over grid-organized data. They are by far the most commonly used type of model when processing images.

A continuous (regression) label does not have a discrete set of possible values, which means possibly an unlimited number of possibilities.

Data vectorization: A process that converts non-numeric data into a numerical format so that it can be used by a machine learning model.

Discrete: A term taken from statistics referring to an outcome taking on only a finite number of values (such as days of the week).

FFNN: The most straightforward way of structuring a neural network, the Feed Forward Neural Network (FFNN) structures neurons in a series of layers, with each neuron in a layer containing weights to all neurons in the previous layer.

Hyperparameters are settings on the model which are not changed during training but can affect how quickly or how reliably the model trains, such as the number of clusters the model should identify.

Log loss is used to calculate how uncertain your model is about the predictions it is generating.

Hyperplane: A mathematical term for a surface that contains more than two planes.

Impute is a common term referring to different statistical tools which can be used to calculate missing values from your dataset.

label refers to data that already contains the solution.

loss function is used to codify the model’s distance from this goal

Machine learning, or ML, is a modern software development technique that enables computers to solve problems by using examples of real-world data.

Model accuracy is the fraction of predictions a model gets right. Discrete: A term taken from statistics referring to an outcome taking on only a finite number of values (such as days of the week). Continuous: Floating-point values with an infinite range of possible values. The opposite of categorical or discrete values, which take on a limited number of possible values.

Model inference is when the trained model is used to generate predictions.

model is an extremely generic program, made specific by the data used to train it.

Model parameters are settings or configurations the training algorithm can update to change how the model behaves.

Model training algorithms work through an interactive process where the current model iteration is analyzed to determine what changes can be made to get closer to the goal. Those changes are made and the iteration continues until the model is evaluated to meet the goals.

Neural networks: a collection of very simple models connected together. These simple models are called neurons. The connections between these models are trainable model parameters called weights.

Outliers are data points that are significantly different from others in the same sample.

Plane: A mathematical term for a flat surface (like a piece of paper) on which two points can be joined by a straight line.

Regression: A common task in supervised machine learning.

In reinforcement learning, the algorithm figures out which actions to take in a situation to maximize a reward (in the form of a number) on the way to reaching a specific goal.

RNN/LSTM: Recurrent Neural Networks (RNN) and the related Long Short-Term Memory (LSTM) model types are structured to effectively represent for loops in traditional computing, collecting state while iterating over some object. They can be used for processing sequences of data.

Silhouette coefficient: A score from -1 to 1 describing the clusters found during modeling. A score near zero indicates overlapping clusters, and scores less than zero indicate data points assigned to incorrect clusters. A

Stop words: A list of words removed by natural language processing tools when building your dataset. There is no single universal list of stop words used by all-natural language processing tools.

In supervised learning, every training sample from the dataset has a corresponding label or output value associated with it. As a result, the algorithm learns to predict labels or output values.

Test dataset: The data withheld from the model during training, which is used to test how well your model will generalize to new data.

Training dataset: The data on which the model will be trained. Most of your data will be here.

Transformer: A more modern replacement for RNN/LSTMs, the transformer architecture enables training over larger datasets involving sequences of data.

In unlabeled data, you don't need to provide the model with any kind of label or solution while the model is being trained.

In unsupervised learning, there are no labels for the training data. A machine learning algorithm tries to learn the underlying patterns or distributions that govern the data.





https://youtu.be/CNNmBtNcccE


Machine learning is synthesizing death metal. It might make your death metal radio DJ nervous – but it could also mean music software works with timbre and time in new ways. That news – plus some comical abuse of neural networks for writing genre-specific lyrics in genres like country – next. Peter Kirn http://cdm.link/2019/04/now-ai-takes-on-writing-death-metal-country-music-hits-more/