A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
Improved EXACT strategy reduces GNN memory consumption and runtime.
Recomputation algorithms collectively refer to a family of methods that aims to reduce the memory consumption of the backpropagation by selectively discarding the intermediate results of the forward propagation and recomputing the discarded results as needed. In this paper, we will propose a novel and efficient recompu…
The paper benchmarks data stream classifiers for human activity recognition on connected devices.
Recent studies on automatic neural architectures search have demonstrated significant performance, competitive to or even better than hand-crafted neural architectures. However, most of the existing network architecture tend to use residual, parallel structures and concatenation block between shallow and deep features …
A method to reduce memory usage in NAS by pruning the search space.
SGQuant reduces GNN memory usage without significant accuracy loss.
Electricity consumption has increased exponentially during the past few decades. This increase is heavily burdening the electricity distributors. Therefore, predicting the future demand for electricity consumption will provide an upper hand to the electricity distributor. Predicting electricity consumption requires man…
A new text classification method using separable convolution reduces memory consumption.
POET enables large neural network training on tiny devices with reduced energy.
TensorOpt finds optimal parallelization strategies for DNN training.
Tail averaging consists in averaging the last examples in a stream. Common techniques either have a memory requirement which grows with the number of samples to average, are not available at every timestep or do not accomodate growing windows. We propose two techniques with a low constant memory cost that perform tail …
Paper proposes a new method to select memory data for online class-incremental learning.
State-of-the-art models are now trained with billions of parameters, reaching hardware limits in terms of memory consumption. This has created a recent demand for memory-efficient optimizers. To this end, we investigate the limits and performance tradeoffs of memory-efficient adaptively preconditioned gradient methods.…
The paper tackles scalability issues in Graph Representation Learning.
In this paper, we compare the performances of FAISS and FENSHSES on nearest neighbor search in Hamming space--a fundamental task with ubiquitous applications in nowadays eCommerce. Comprehensive evaluations are made in terms of indexing speed, search latency and RAM consumption. This comparison is conducted towards a b…
Training deep learning models on mobile devices recently becomes possible, because of increasing computation power on mobile hardware and the advantages of enabling high user experiences. Most of the existing work on machine learning at mobile devices is focused on the inference of deep learning models (particularly co…
This paper discusses how usage patterns and preferences of inhabitants can be learned efficiently to allow smart homes to autonomously achieve energy savings. We propose a frequent sequential pattern mining algorithm suitable for real-life smart home event data. The performance of the proposed algorithm is compared to …
SDA method reduces memory and time for assimilating noisy geophysical data.
SliceOut speeds up deep learning training without sacrificing accuracy.
Embedding representations power machine intelligence in many applications, including recommendation systems, but they are space intensive -- potentially occupying hundreds of gigabytes in large-scale settings. To help manage this outsized memory consumption, we explore mixed dimension embeddings, an embedding layer arc…
CICLAD efficiently mines frequent closed itemsets from data streams with minimal memory usage.
HomoODE connects DEQs and Neural ODEs via homotopy continuation, improving accuracy and memory efficiency.
Model quantization is leveraged to reduce the memory consumption and the computation time of deep neural networks. This is achieved by representing weights and activations with a lower bit resolution when compared to their high precision floating point counterparts. The suitable level of quantization is directly relate…
Kernel methods have great promise for learning rich statistical representations of large modern datasets. However, compared to neural networks, kernel methods have been perceived as lacking in scalability and flexibility. We introduce a family of fast, flexible, lightly parametrized and general purpose kernel learning …
Most electricity systems worldwide are deploying advanced metering infrastructures to collect relevant operational data. In particular, smart meters allow tracking electricity load consumption at a very disaggregated level and at high frequency rates. This data opens the possibility of developing new forecasting models…
DPZero fine-tunes large models privately without backpropagation.
This paper reviews methods to create compact neural networks for IoT applications.
Deep neural networks (DNNs) depend on the storage of a large number of parameters, which consumes an important portion of the energy used during inference. This paper considers the case where the energy usage of memory elements can be reduced at the cost of reduced reliability. A training algorithm is proposed to optim…
We present here the Temporal Clustering Algorithm (TCA), an incremental learning algorithm applicable to problems of anticipatory computing in the context of the Internet of Things. This algorithm was tested in a specific prediction scenario of consumption of an electric water dispenser typically used in tropical count…
This paper proposes a hardware-oriented dropout algorithm, which is efficient for field programmable gate array (FPGA) implementation. In deep neural networks (DNNs), overfitting occurs when networks are overtrained and adapt too well to training data. Consequently, they fail in predicting unseen data used as test data…
TASO optimizes CNN models for memory-constrained devices.
Neural architecture search (NAS) has a great impact by automatically designing effective neural network architectures. However, the prohibitive computational demand of conventional NAS algorithms (e.g. GPU hours) makes it difficult to \emph{directly} search the architectures on large-scale tasks (e.g. ImageNet).…
Time series forecasting is an important problem across many domains, including predictions of solar plant energy output, electricity consumption, and traffic jam situation. In this paper, we propose to tackle such forecasting problem with Transformer [1]. Although impressed by its performance in our preliminary study, …
Deep neural networks solve stochastic control problems with delay.
Paper improves GNN inference speed and memory usage.
RSmote improves PINNs accuracy with less memory usage.
Training a neural network using backpropagation algorithm requires passing error gradients sequentially through the network. The backward locking prevents us from updating network layers in parallel and fully leveraging the computing resources. Recently, there are several works trying to decouple and parallelize the ba…
A new deep learning method using Boolean logic reduces training and inference energy.
Many key algorithms in 3-manifold topology involve the enumeration of normal surfaces, which is based upon the double description method for finding the vertices of a convex polytope. Typically we are only interested in a small subset of these vertices, thus opening the way for substantial optimization. Here we give an…
The Long-Short-Term-Memory Recurrent Neural Networks (LSTM RNNs) are a popular class of machine learning models for analyzing sequential data. Their training on modern GPUs, however, is limited by the GPU memory capacity. Our profiling results of the LSTM RNN-based Neural Machine Translation (NMT) model reveal that fea…
We present a new approach to modeling sequential data: the deep equilibrium model (DEQ). Motivated by an observation that the hidden layers of many existing deep sequence models converge towards some fixed point, we propose the DEQ approach that directly finds these equilibrium points via root-finding. Such a method is…
Airlines optimize fuel loading with better flight time predictions.
Computer vision performances have been significantly improved in recent years by Convolutional Neural Networks(CNN). Currently, applications using CNN algorithms are deployed mainly on general purpose hardwares, such as CPUs, GPUs or FPGAs. However, power consumption, speed, accuracy, memory footprint, and die size sho…
Memory-augmented neural networks (MANNs) refer to a class of neural network models equipped with external memory (such as neural Turing machines and memory networks). These neural networks outperform conventional recurrent neural networks (RNNs) in terms of learning long-term dependency, allowing them to solve intrigui…
VPFD uses vocoder features for adversarial training in VC.
Recurrent neural networks (RNNs) have shown excellent performance in processing sequence data. However, they are both complex and memory intensive due to their recursive nature. These limitations make RNNs difficult to embed on mobile devices requiring real-time processes with limited hardware resources. To address the…
Study optimal retirement time and consumption with habitual persistence.
Thermal dynamics modeling has been a critical issue in building heating, ventilation, and air-conditioning (HVAC) systems, which can significantly affect the control and maintenance strategies. Due to the uniqueness of each specific building, traditional thermal dynamics modeling approaches heavily depending on physics…