<i>p</i>-Adic statistical field theory and convolutional deep Boltzmann machines

نویسندگان

چکیده

Understanding how deep learning architectures work is a central scientific problem. Recently, correspondence between neural networks (NNs) and Euclidean quantum field theories (QFTs) has been proposed. This investigates this in the framework of p-adic statistical (SFTs) (NNs). In case, fields are real-valued functions defined on an infinite regular rooted tree with valence p, fixed prime number. provides topology for continuous Boltzmann machine (DBM), which identified theory (SFT) tree. framework, there natural method to discretize SFTs. Each discrete SFT corresponds (BM) tree-like topology. allows us recover standard DBMs gives new convolutional DBMs. The use O(N) parameters while classical ones O(N^{2}) parameters.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Deep Boltzmann Machines

We present a new learning algorithm for Boltzmann machines that contain many layers of hidden variables. Data-dependent expectations are estimated using a variational approximation that tends to focus on a single mode, and dataindependent expectations are approximated using persistent Markov chains. The use of two quite different techniques for estimating the two types of expectation that enter...

متن کامل

Convolutional Restricted Boltzmann Machines for Feature Learning

In this thesis, we present a method for learning problem-specific hierarchical features specialized for vision applications. Recently, a greedy layerwise learning mechanism has been proposed for tuning parameters of fully connected hierarchical networks. This approach views layers of a network as Restricted Boltzmann Machines (RBM), and trains them separately from the bottom layer upwards. We d...

متن کامل

On Training Deep Boltzmann Machines

The deep Boltzmann machine (DBM) has been an important development in the quest for powerful “deep” probabilistic models. To date, simultaneous or joint training of all layers of the DBM has been largely unsuccessful with existing training methods. We introduce a simple regularization scheme that encourages the weight vectors associated with each hidden unit to have similar norms. We demonstrat...

متن کامل

Multi-Prediction Deep Boltzmann Machines

We introduce the multi-prediction deep Boltzmann machine (MP-DBM). The MPDBM can be seen as a single probabilistic model trained to maximize a variational approximation to the generalized pseudolikelihood, or as a family of recurrent nets that share parameters and approximately solve different inference problems. Prior methods of training DBMs either do not perform well on classification tasks ...

متن کامل

Deep Boltzmann Machines and the Centering Trick

Deep Boltzmann machines are in theory capable of learning efficient representations of seemingly complex data. Designing an algorithm that effectively learns the data representation can be subject to multiple difficulties. In this chapter, we present the “centering trick” that consists of rewriting the energy of the system as a function of centered states. The centering trick improves the condi...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Progress of theoretical and experimental physics

سال: 2023

ISSN: ['1347-4081', '0033-068X']

DOI: https://doi.org/10.1093/ptep/ptad061