charlesmartin14.wordpress.com
Machine Learning with Missing Labels Part 3: Experiments | Machine Learning
https://charlesmartin14.wordpress.com/2014/11/01/machine-learning-with-missing-labels-part-3-experiments
Notes, thoughts, and practice of applied machine learning. Machine Learning with Missing Labels Part 3: Experiments. Machine Learning with Missing Labels Part 3: Experiments. November 1, 2014. January 4, 2016. Charles H Martin, PhD. In this series of posts we look at Transductive and SemiSupervised Learning–an old problem, a hard problem, and a fundamental problem Machine Learning. Unlike Deep Learning or large scale ML,. We want to learn as much as we can from as labeled little data as possible. Have re...
charlesmartin14.wordpress.com
Machine Learning with Missing Labels Part 2: The UniverSVM | Machine Learning
https://charlesmartin14.wordpress.com/2014/09/30/machine-learning-with-missing-labels-part-2-advanced-svms
Notes, thoughts, and practice of applied machine learning. Machine Learning with Missing Labels Part 2: The UniverSVM. Machine Learning with Missing Labels Part 2: The UniverSVM. September 30, 2014. December 1, 2014. Charles H Martin, PhD. Ever wonder what Google DeepMind. They just released a paper on Semi-Supervised learning with Deep Generative Models. What is Semi Supervised Learning (SSL)? In this series of posts, we go back to basics and take a look. 8211;ala Vapnik 2006. Plus, there is code! In th...
charlesmartin14.wordpress.com
Machine Learning with Missing Labels: Transductive SVMs | Machine Learning
https://charlesmartin14.wordpress.com/2014/09/23/machine-learning-with-missing-labels-transductive-svms
Notes, thoughts, and practice of applied machine learning. Machine Learning with Missing Labels: Transductive SVMs. Machine Learning with Missing Labels: Transductive SVMs. September 23, 2014. February 20, 2015. Charles H Martin, PhD. SVMs are great for building text classifiers–if you have a set of very high quality, labeled documents. Frequently, we just don’t have enough labelled data! What can we do? Pay mechanical Turks to label the documents. This is actually harder than it sounds. Since the initia...
charlesmartin14.wordpress.com
Kernels Part 1: What is an RBF Kernel? Really? | Machine Learning
https://charlesmartin14.wordpress.com/2012/02/06/kernels_part_1
Notes, thoughts, and practice of applied machine learning. Kernels Part 1: What is an RBF Kernel? Kernels Part 1: What is an RBF Kernel? February 6, 2012. April 12, 2015. Charles H Martin, PhD. My first blog on machine learning is to discuss a pet peeve I have about working in the industry, namely why not to apply an RBF kernel to text classification tasks. I wrote this as a follow up to a Quora Answer on the subject:. Smola, Scholkopf, and Muller. I expand on one point–. Documents / training instances (.
charlesmartin14.wordpress.com
Noisy Time Series II: Earth Quakes, Black Holes, and Machine Learning | Machine Learning
https://charlesmartin14.wordpress.com/2012/11/07/noisy-time-series-ii
Notes, thoughts, and practice of applied machine learning. Noisy Time Series II: Earth Quakes, Black Holes, and Machine Learning. Noisy Time Series II: Earth Quakes, Black Holes, and Machine Learning. November 7, 2012. October 3, 2013. Charles H Martin, PhD. Recently , 7 Italian Scientists have been sentenced in prison for manslaughter for failing to predict an Earthquake in 2009. So how in the world would a Machine Learning Scientist predict an Earthquake? Eh…not so much! Scale Invariance in Nature:.
charlesmartin14.wordpress.com
Advances in Convex NMF: Linear Programming | Machine Learning
https://charlesmartin14.wordpress.com/2013/05/06/advances-in-convex-nmf-part-1-linear-programming
Notes, thoughts, and practice of applied machine learning. Advances in Convex NMF: Linear Programming. Advances in Convex NMF: Linear Programming. May 6, 2013. March 24, 2014. Charles H Martin, PhD. Today I am going to look at a very important advance in one of my favorite Machine Learning algorithms, NMF (Non-Negative Matrix Factorization). Several approaches exist to improve traditional NMF [1], such as sparse NMF. Jordan et. al. [5]. In fact, it has been known for some time that CNMF could be re-formu...
charlesmartin14.wordpress.com
Spectral Clustering: A quick overview | Machine Learning
https://charlesmartin14.wordpress.com/2012/10/09/spectral-clustering
Notes, thoughts, and practice of applied machine learning. Spectral Clustering: A quick overview. Spectral Clustering: A quick overview. October 9, 2012. March 10, 2015. Charles H Martin, PhD. Here, I give a brief tutorial on the theory of Spectral Clustering and how it is implemented in open source packaages. At some point I will rewrite some of this and add a review of this recent paper Robust and Scalable Graph-Based Semisupervised Learning. Spectral (or Subspace) Clustering. Project your data into.
charlesmartin14.wordpress.com
Metric Learning: Some Quantum Statistical Mechanics | Machine Learning
https://charlesmartin14.wordpress.com/2013/11/14/metric-learning-some-quantum-statistical-mechanics
Notes, thoughts, and practice of applied machine learning. Metric Learning: Some Quantum Statistical Mechanics. Metric Learning: Some Quantum Statistical Mechanics. November 14, 2013. July 7, 2014. Charles H Martin, PhD. I wrote this for my buddy Sebass; just a quick review of quantum stat mech:. In our last post, we presented a formalism for music recommendations called Logistic Markov Embedding (LME). We will follow the classic text by T. Hill. Probabilities and the Partition Functions. Frequently we s...
charlesmartin14.wordpress.com
Convex Relaxations of Transductive Learning | Machine Learning
https://charlesmartin14.wordpress.com/2015/03/14/convex-relaxations-of-transductive-learning
Notes, thoughts, and practice of applied machine learning. Convex Relaxations of Transductive Learning. Convex Relaxations of Transductive Learning. March 14, 2015. July 31, 2016. Charles H Martin, PhD. Why are SVMs interesting? It is just a better way to do Logistic Regression? Is it the Kernel Trick? And does this even matter now that Deep Learning is everywhere? Historically, convex optimization was seen as the path to central planing an entire economy. A great new book, Red Plenty. Today there are Tr...
charlesmartin14.wordpress.com
Why does Deep Learning work? | Machine Learning
https://charlesmartin14.wordpress.com/2015/03/25/why-does-deep-learning-work
Notes, thoughts, and practice of applied machine learning. Why does Deep Learning work? Why does Deep Learning work? March 25, 2015. July 13, 2015. Charles H Martin, PhD. Why does Deep Learning work? This is the big question on everyone’s mind these days. C’mon we all know the answer already:. 8220;the long-term behavior of certain neural network models are governed by the statistical mechanism of infinite-range Ising spin-glass Hamiltonians” [1]. In other words,. Multilayer Neural Networks are just.