arXiv subjects

Nicolae Tapus

Publications and source records attributed to Nicolae Tapus.

17 recordsLinked to original sources

arXiv2021

The Accuracy and Efficiency of Posit Arithmetic

Motivated by the increasing interest in the posit numeric format, in this paper we evaluate the accuracy and efficiency of posit arithmetic in contrast to the traditional IEEE 754 32-bit floating-point (FP32) arithmetic. We first design and implement a Posit Arithmetic Unit (PAU), called POSAR, with flexible bit-sized arithmetic suitable for applications that can trade accuracy for savings in chip area. Next, we analyze the accuracy and efficiency of POSAR with a series of benchmarks including mathematical computations, ML kernels, NAS Parallel Benchmarks (NPB), and Cifar-10 CNN. This analysis is done on our implementation of POSAR integrated into a RISC-V Rocket Chip core in comparison with the IEEE 754-based Floting Point Unit (FPU) of Rocket Chip. Our analysis shows that POSAR can outperform the FPU, but the results are not spectacular. For NPB, 32-bit posit achieves better accuracy than FP32 and improves the execution by up to 2%. However, POSAR with 32-bit posit needs 30% more FPGA resources compared to the FPU. For classic ML algorithms, we find that 8-bit posits are not suitable to replace FP32 because they exhibit low accuracy leading to wrong results. Instead, 16-bit posit offers the best option in terms of accuracy and efficiency. For example, 16-bit posit achieves the same Top-1 accuracy as FP32 on a Cifar-10 CNN with a speedup of 18%.

cs.AR↗

arXiv2020

CloudifierNet -- Deep Vision Models for Artificial Image Processing

Today, more and more, it is necessary that most applications and documents developed in previous or current technologies to be accessible online on cloud-based infrastructures. That is why the migration of legacy systems including their hosts of documents to new technologies and online infrastructures, using modern Artificial Intelligence techniques, is absolutely necessary. With the advancement of Artificial Intelligence and Deep Learning with its multitude of applications, a new area of research is emerging - that of automated systems development and maintenance. The underlying work objective that led to this paper aims to research and develop truly intelligent systems able to analyze user interfaces from various sources and generate real and usable inferences ranging from architecture analysis to actual code generation. One key element of such systems is that of artificial scene detection and analysis based on deep learning computer vision systems. Computer vision models and particularly deep directed acyclic graphs based on convolutional modules are generally constructed and trained based on natural images datasets. Due to this fact, the models will develop during the training process natural image feature detectors apart from the base graph modules that will learn basic primitive features. In the current paper, we will present the base principles of a deep neural pipeline for computer vision applied to artificial scenes (scenes generated by user interfaces or similar). Finally, we will present the conclusions based on experimental development and benchmarking against state-of-the-art transfer-learning implemented deep vision models.

cs.CV↗

arXiv2019

Deep recommender engine based on efficient product embeddings neural pipeline

Predictive analytics systems are currently one of the most important areas of research and development within the Artificial Intelligence domain and particularly in Machine Learning. One of the "holy grails" of predictive analytics is the research and development of the "perfect" recommendation system. In our paper, we propose an advanced pipeline model for the multi-task objective of determining product complementarity, similarity and sales prediction using deep neural models applied to big-data sequential transaction systems. Our highly parallelized hybrid model pipeline consists of both unsupervised and supervised models, used for the objectives of generating semantic product embeddings and predicting sales, respectively. Our experimentation and benchmarking processes have been done using pharma industry retail real-life transactional Big-Data streams.

cs.IR↗

arXiv2019

Advanced Customer Activity Prediction based on Deep Hierarchic Encoder-Decoders

Product recommender systems and customer profiling techniques have always been a priority in online retail. Recent machine learning research advances and also wide availability of massive parallel numerical computing has enabled various approaches and directions of recommender systems advancement. Worth to mention is the fact that in past years multiple traditional "offline" retail business are gearing more and more towards employing inferential and even predictive analytics both to stock-related problems such as predictive replenishment but also to enrich customer interaction experience. One of the most important areas of recommender systems research and development is that of Deep Learning based models which employ representational learning to model consumer behavioral patterns. Current state of the art in Deep Learning based recommender systems uses multiple approaches ranging from already classical methods such as the ones based on learning product representation vector, to recurrent analysis of customer transactional time-series and up to generative models based on adversarial training. Each of these methods has multiple advantages and inherent weaknesses such as inability of understanding the actual user-journey, ability to propose only single product recommendation or top-k product recommendations without prediction of actual next-best-offer. In our work we will present a new and innovative architectural approach of applying state-of-the-art hierarchical multi-module encoder-decoder architecture in order to solve several of current state-of-the-art recommender systems issues. Our approach will also produce by-products such as product need-based segmentation and customer behavioral segmentation - all in an end-to-end trainable approach. Finally, we will present a couple methods that solve known retail & distribution pain-points based on the proposed architecture.

cs.IR↗

arXiv2013

Efficient Online Algorithmic Strategies for Several Two-Player Games with Different or Identical Player Roles

In this paper we introduce novel algorithmic strategies for effciently playing two-player games in which the players have different or identical player roles. In the case of identical roles, the players compete for the same objective (that of winning the game). The case with different player roles assumes that one of the players asks questions in order to identify a secret pattern and the other one answers them. The purpose of the first player is to ask as few questions as possible (or that the questions and their number satisfy some previously known constraints) and the purpose of the secret player is to answer the questions in a way that will maximize the number of questions asked by the first player (or in a way which forces the first player to break the constraints of the game). We consider both previously known games (or extensions of theirs) and new types of games, introduced in this paper.

cs.GT↗

arXiv2011

A distributed service for on demand end to end optical circuits

In this paper we present a system for monitoring and controlling dynamic network circuits inside the USLHCNet network. This distributed service system provides in near real-time complete topological information for all the circuits, resource allocation and usage, accounting, detects automatically failures in the links and network equipment, generate alarms and has the functionality to take automatic actions. The system is developed based on the MonALISA framework, which provides a robust monitoring and controlling service oriented architecture, with no single points of failure.

cs.DC↗

arXiv2010

Brief Announcement: Decentralized Construction of Multicast Trees Embedded into P2P Overlay Networks based on Virtual Geometric Coordinates

In this paper we consider the problem of efficiently constructing in a fully distributed manner multicast trees which are embedded into P2P overlays using virtual geometric node coordinates. We consider two objectives: to minimize the number of messages required for constructing a multicast tree by using the geometric properties of the P2P overlay, and to construct stable multicast trees when the lifetime durations of the peers are known.

cs.DC↗

arXiv2010

Algorithmic Solutions for Several Offline Constrained Resource Processing and Data Transfer Multicriteria Optimization Problems

In this paper we present novel algorithmic solutions for several resource processing and data transfer multicriteria optimization problems. The results of most of the presented techniques are strategies which solve the considered problems (almost) optimally. Thus, the developed algorithms construct intelligent strategies which can be implemented by agents in specific situations. All the described solutions make use of the properties of the considered problems and, thus, they are not applicable to a very general class of problems. However, by considering the specific details of each problem, we were able to obtain very efficient results.

cs.DS↗

arXiv2010

Practical Range Aggregation, Selection and Set Maintenance Techniques

In this paper we present several new and very practical methods and techniques for range aggregation and selection problems in multidimensional data structures and other types of sets of values. We also present some new extensions and applications for some fundamental set maintenance problems.

cs.DS↗

arXiv2010

Efficient Upload Bandwidth Estimation and Communication Resource Allocation Techniques

In this paper we address two problems, for which we present novel, efficient, algorithmic solutions. The first problem is motivated by practical situations and is concerned with the efficient estimation of the upload bandwidth of a machine, particularly in the context of a peer-to-peer content sharing and distribution application. The second problem is more of a theoretical nature and considers a constrained communication resource allocation situation.

cs.NI↗

arXiv2009

New Algorithmic Approaches for Computing Optimal Network Paths with Several Types of QoS Constraints

The problem of efficiently delivering data within networks is very important nowadays, especially in the context of the large volumes of data which are being produced each year and of the increased data access needs of the users. Efficient data delivery strategies must satisfy several types of Quality of Service (QoS) constraints which are imposed by the data consumers. One possibility of achieving this goal (particularly in the case of in-order data transfers) is to choose a satisfactory network delivery path. In this paper we present novel algorithmic approaches for computing optimal network paths which satisfy several types of (QoS) constraints.

cs.DS↗

arXiv2009

Practical Algorithmic Techniques for Several String Processing Problems

The domains of data mining and knowledge discovery make use of large amounts of textual data, which need to be handled efficiently. Specific problems, like finding the maximum weight ordered common subset of a set of ordered sets or searching for specific patterns within texts, occur frequently in this context. In this paper we present several novel and practical algorithmic techniques for processing textual data (strings) in order to efficiently solve multiple problems. Our techniques make use of efficient string algorithms and data structures, like KMP, suffix arrays, tries and deterministic finite automata.

cs.DS↗

arXiv2009

Efficient Offline Algorithmic Techniques for Several Packet Routing Problems in Distributed Systems

In this paper we consider several problems concerning packet routing in distributed systems. Each problem is formulated using terms from Graph Theory and for each problem we present efficient, novel, algorithmic techniques for computing optimal solutions. We address topics like: bottleneck paths (trees), optimal paths with non-linear costs, optimal paths with multiple optimization objectives, maintaining aggregate connectivity information under a sequence of network link failures, and several others.

cs.DS↗

arXiv2009

Towards a Centralized Scheduling Framework for Communication Flows in Distributed Systems

The overall performance of a distributed system is highly dependent on the communication efficiency of the system. Although network resources (links, bandwidth) are becoming increasingly more available, the communication performance of data transfers involving large volumes of data does not necessarily improve at the same rate. This is due to the inefficient usage of the available network resources. A solution to this problem consists of data transfer scheduling techniques, which manage and allocate the network resources in an efficient manner. In this paper we present several online and offline data transfer optimization techniques, in the context of a centrally controlled distributed system.

cs.DS↗

arXiv2009

Offline Algorithms for Several Network Design, Clustering and QoS Optimization Problems

In this paper we address several network design, clustering and Quality of Service (QoS) optimization problems and present novel, efficient, offline algorithms which compute optimal or near-optimal solutions. The QoS optimization problems consist of reliability improvement (by computing backup shortest paths) and network link upgrades (in order to reduce the latency on several paths). The network design problems consist of determining small diameter networks, as well as very well connected and regular network topologies. The network clustering problems consider only the restricted model of static and mobile path networks, for which we were able to develop optimal algorithms.

cs.DS↗

arXiv2009

Data Distribution Optimization using Offline Algorithms and a Peer-to-Peer Small Diameter Tree Architecture with Bounded Node Degrees

Multicast data transfers occur in many distributed systems and applications (e.g. IPTV, Grids, content delivery networks). Because of this, efficient multicast data distribution optimization techniques are required. In the first part of this paper we present a small diameter, bounded degree, collaborative peer-to-peer multicast tree architecture, which supports dynamic node arrivals and departures making local decisions only. The architecture is fault tolerant and, at low arrival and departure rates, converges towards a theoretically optimal structure. In the second part of the paper we consider several offline data distribution optimization problems, for which we present novel and time-efficient algorithmic solutions.

cs.DC↗

arXiv2009

Offline Algorithmic Techniques for Several Content Delivery Problems in Some Restricted Types of Distributed Systems

In this paper we consider several content delivery problems (broadcast and multicast, in particular) in some restricted types of distributed systems (e.g. optical Grids and wireless sensor networks with tree-like topologies). For each problem we provide efficient algorithmic techniques for computing optimal content delivery strategies. The techniques we present are offline, which means that they can be used only when full information is available and the problem parameters do not fluctuate too much.

cs.DS↗