added an overview page and some pages to be filled in later

content/models/overview.md

@@ -0,0 +1,60 @@
+# An Overview of Machine Learning Methods in High Energy Physics
+
+## Introduction
+
+In the realm of high energy physics, the sheer volume and complexity of data generated by experiments like those at the Compact Muon Solenoid (CMS) at CERN, necessitate the use of sophisticated data analysis techniques. Machine Learning (ML), with its ability to learn patterns and make predictions from large datasets, has emerged as a powerful tool in this field. This article provides an overview of various machine learning methods and models used in high energy physics, aimed at physicists working with CMS who are interested in understanding and potentially applying ML techniques.
+
+## Supervised Learning
+
+Supervised learning is a type of ML where the model is trained on a labeled dataset, learning to predict the output from the input data. In high energy physics, this can be used to classify events or particles based on their properties.
+
+1. **Decision Trees and Random Forests**: Decision trees are simple yet powerful models used for classification and regression tasks. They are popular due to their interpretability - the model's decisions can be visualized and understood easily. Random Forests, an ensemble of decision trees, improve prediction accuracy by averaging the predictions of multiple decision trees trained on different subsets of the data.
+
+2. **Support Vector Machines (SVMs)**: SVMs are effective in high-dimensional spaces, making them suitable for problems with complex domains. They work by finding the hyperplane that best separates the classes in the data.
+
+3. **Neural Networks**: These algorithms, modeled loosely after the human brain, are designed to recognize patterns. They are used for complex tasks like particle identification and event reconstruction.
+
+## Unsupervised Learning
+
+Unsupervised learning is used when the information used to train is neither classified nor labeled. This method can infer a function to describe a hidden structure from unlabeled data.
+
+1. **Clustering Algorithms**: These algorithms group a set of objects in such a way that objects in the same group are more similar to each other than to those in other groups. This can be useful in high energy physics for grouping similar events or particles together.
+
+2. **Principal Component Analysis (PCA)**: PCA reduces the dimensionality of the data by transforming it to a new set of variables, the principal components, which are uncorrelated. This can be useful for visualizing high-dimensional data or for preprocessing before applying 
+other ML methods.
+
+3. **Autoencoders**: Autoencoders are neural networks that learn to efficiently encode and decode input data. They compress data into a lower-dimensional form and reconstruct it back, minimizing the difference between the original and reconstructed data. In high energy physics, they can be used for noise reduction and anomaly detection by learning a representation of "normal" data and identifying deviations from it.
+
+## Reinforcement Learning
+
+Reinforcement learning is a type of ML where an agent learns to behave in an environment by performing actions and observing the results. While not as commonly used in high energy physics as supervised and unsupervised learning, reinforcement learning has potential applications in areas like optimizing experimental settings.
+
+## Deep Learning
+
+Deep learning, a subset of machine learning, uses neural networks with many layers ("deep" structures) to learn from large amounts of data. Deep learning models are widely used in high energy physics for tasks such as particle identification, event reconstruction, and event classification.
+
+1. **Convolutional Neural Networks (CNNs)**: CNNs are particularly effective for image data, making them useful for analyzing rectelinear data from detectors.
+
+2. **Recurrent Neural Networks (RNNs)**: RNNs are designed for sequential data, and could be used for analyzing time-series data from experiments.
+
+3. **Graph Neural Networks (GNNs)**: GNNs are best utilized on data that can be represented as a graph. They are considered a generalization of CNNs.
+
+Generative Models
+
+Generative models, such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs), are used to generate new data instances that resemble the training data. In high energy physics, these models can be used for simulating high-dimensional distributions of particle interactions.
+
+## Anomaly Detection
+
+Anomaly detection is the identification of rare items, events, or observations which raise suspicions by differing significantly from the majority of the data. In the context of high energy physics, anomaly detection can be used to identify new particles or unexpected behavior in data from particle accelerators.
+
+## Model Disambiguation
+
+* [Autoencoders](../training/autoencoders.md)
+* [CNNs](CNNs.md)
+* [RNNs](RNNs.md)
+* [GNNs](GNNs.md)
+* [Decision Trees](decision_trees.md)
+* [Perceptrons](perceptrons.md)
+
+[TODO: add more]
+

mkdocs.yml

@@ -163,5 +163,8 @@ nav:
     - Training:
       - Training as a Service:
         - MLaaS4HEP: training/MLaaS4HEP.md
+    - Models and Usecases:
+      - Overview: models/overview.md
       - Autoencoders: training/autoencoders.md
+
 #    - Benchmarking: