Research and applications in Machine Learning are limited by computational resources, while 1% of the world's electricity goes into calculating 34 billion billion SHA-256 hashes per second, four orders of magnitude more than the 200 petaflop power of the world's most powerful supercomputer. The work presented here describes how a simple soft fork on Bitcoin can adapt these incomparable resources to a global distributed computer. By creating an infrastructure and ledger fully compatible with blockchain technology, the hashes can be replaced with stochastic optimizations such as Deep Net training, inverse problems such as GANs, and arbitrary NP computations.
Amrutanshu Panigrahi, Ajit Kumar Nayak, Rourab Paul
Blockchain technology facilitates transparency, de-centralization, immutability, and security for each transaction. The decentralization characteristics enable the block chain network to have different nodes as the Certificate Authority (CA) for a different transactions. Choosing a different validator for every individual transaction may increase the security perspective of the transaction, but it can lead to one major concern such as the network overhead increase. There can be several numbers of transactions for a single network and selecting a validator every time can cause a block propagation delay which also decreases the network efficiency. To avoid such kind of issue clustering of network nodes can be an emerging solution. In the current research work, the entire network participating nodes are considered for making different clusters based on the response time factor. Every cluster has the average response time as the threshold value. The node that wants to initiate a transaction needs to raise a minimum response time requirement and the cluster selection procedure will be executed based on the received response time value. In the current research work, the K-Means clustering with the Elbow method as an internal validation method is considered to decide the number of clusters. The computational time for both cases is compared to measure the effectiveness of the clustering process on the blockchain network.
In blockchain (BC) systems, smart contracts (SCs) often use a transaction-driven execution model. However, real-world financial applications are complex, and often driven by events instead. This paper presents an intelligent event-driven framework to automatically trigger SC execution by collecting event information, and classifying these events, and analyzing various relationship among these events, and use a service-model for SCs to register interested events so that when the concerned event happened, they can be automatically triggered. This paper uses oracle machines (OMs) to collect event information, EEG (Event Evolution Graph) to analyze relationship among events. This paper then uses multi-layer perceptron model to evaluate the design. This paper proposes a new method to solve the scalability and performance challenges in the SC ecosystem, which shows that the event-driven execution model would be more intelligent.
This work aims to reduce transaction time by integrating the mining process with the sending of transactions to reduce overall transaction time. This work also decreased the overhead of mining in a private network, by reducing the production of empty blocks in the network which saves energy, storage space, network bandwidth and computational complexity. We proposed an Auto Integrated Mining (AIM) algorithm which starts the mining process only when there is at least one pending transaction in the network, and stops mining as soon as the pending transactions are mined. The results show that the AIM algorithm reduced the number of mined blocks in a 12 hour period by producing only 24% of the original number of blocks. The proposed algorithm is also able to reduce the storage used to save chaindata by 16%. The experiment shows that the mining latency of AIM varied between 200-650ms when the number of pending transactions was between 1-1,000, and had a latency between 350-1,350ms when there were 1,000-10,000 pending transactions in the private blockchain network.
Olayemi Mikail Olaniyi, Abraham Ayegba Alfa, Buhari Ugbede Umar
Blockchain technology has gained lots of traction in the past five years due to the innovations introduced in digital currency, the Bitcoin. This technology is powered by distributed ledger technology, which is a distributed database system. It is often renowned for decentralization, anti-attack, and unfalsified attributes making it a top choice in several non-monetary applications. In fact, the problem of privacy and security of the Internet of Things has been undertaken aggressively with Blockchain. Several problems have been identified with blockchain technology such as large delays and lack of support for real-time transaction processing, authorization, node verification, and consensus mechanisms. This article intends to provide a comprehensive survey on the recent advances and solutions to the problems of blockchain technology by leveraging the artificial intelligence approaches. The outcomes of this study will provide valuable information and guidance on the design of Blockchain-based systems to support time-sensitive and real-time specific applications and processes.
Michael Pacheco, Gustavo A. Oliva, Gopi Krishnan Rajbahadur, Ahmed E. Hassan
The Ethereum platform allows developers to implement and deploy applications called Dapps onto the blockchain for public use through the use of smart contracts. To execute code within a smart contract, a paid transaction must be issued towards one of the functions that are exposed in the interface of a contract. However, such a transaction is only processed once one of the miners in the peer-to-peer network selects it, adds it to a block, and appends that block to the blockchain This creates a delay between transaction submission and code execution. It is crucial for Dapp developers to be able to precisely estimate when transactions will be processed, since this allows them to define and provide a certain Quality of Service (QoS) level (e.g., 95% of the transactions processed within 1 minute). However, the impact that different factors have on these times have not yet been studied. Processing time estimation services are used by Dapp developers to achieve predefined QoS. Yet, these services offer minimal insights into what factors impact processing times. Considering the vast amount of data that surrounds the Ethereum blockchain, changes in processing times are hard for Dapp developers to predict, making it difficult to maintain said QoS. In our study, we build random forest models to understand the factors that are associated with transaction processing times. We engineer several features that capture blockchain internal factors, as well as gas pricing behaviors of transaction issuers. By interpreting our models, we conclude that features surrounding gas pricing behaviors are very strongly associated with transaction processing times. Based on our empirical results, we provide Dapp developers with concrete insights that can help them provide and maintain high levels of QoS.
At present, water resource information management in China is mainly a centralized model, and there exist some problems such as high cost, low efficiency, and data storage insecurity. Blockchain technology provides a good solution which can create an efficient trust mechanism among the links in the process of water resource utilization. It guarantees the security of the data, avoiding the sudden collapse of the central institutions caused by some normal operations of the entire system. Based on a decentralization blockchain, we propose a decentralized water resource information management system for the whole process of âsupply-use-consumption-discharge,â which improves the traditional water data storage. Specifically, the monitoring and business data are encrypted by the blockchain and are transmitted using a peer-to-peer network. Moreover, the centralized management mode is changed and part of the management work is dispersed to each node. Thus, decisions and measures can be made and implemented quickly after discovering problems to improve the efficiency of information transmission and management. In addition, two typical blockchain-based application scenarios for water resource management are designed. A blockchain-based approach makes issuing and monitoring water abstraction permits more convenient and obtaining license information more secure and verifiable. A reliable mechanism for tracing water quality ensures the accuracy and reliability of water quality information, enables the detection of locations with inadequate water quality, and clarifies peopleâs responsibility, thus guaranteeing the water safety of the residents.
Roi Bar-Zur, Ameer Abu-Hanna, Ittay Eyal, Aviv Tamar
Blockchain technology is responsible for the emergence of cryptocurrencies, such as Bitcoin and Ethereum. The security of a blockchain protocol relies on the incentives of its participants. Selfish mining is a form of deviation from the protocol where a participant can gain more than her fair share. Previous analyses of selfish mining make easing, non-realistic assumptions. We introduce a more realistic model with varying block rewards in the form of transaction fees. However, this comes at the cost of an intractable state space. To solve the complex model, we introduce WeRLman, a novel method based on deep Reinforcement Learning (deep RL). Using WeRLman, we show reward variability can significantly hurt blockchain security.
Social media and financial markets are ecosystems bound to interact and overlap more and more. For example, cryptocurrencies, Bitcoin in the first place, have long been among the hottest topics on social media. Here we illustrate a methodology for correlating Bitcoin price trends and social media that operates on the aggregation of news from multiple social networks as typically occurs on dedicated channels in sites such as Reddit. For this purpose, we define general laws to map the financial fluctuations of Bitcoin in a space defined on three dimensions: content volume, sentiment and time. These laws provide the foundation upon which to build effective methodologies for social media-based prediction of Bitcoin performance.
Abstract Tracking unseen water in products (Embedded Virtual Water) has generated great interest in the scientific community. This water transfers between geographies via suppliers, manufacturers, distributors, retailers and customers in multiple phases. However, the Virtual Water Trading System lacks proper accounting standards, established protocols and processes in the context of product manufacturing. Therefore, there is a need to establish a technology platform to handle the complex virtual water international trade. Such a platform should uphold transparency and create âwater consciousnessâ and awareness among companies and consumers. The concept of a virtual water currency and blockchain technology platform together can manage these processes. Blockchain helps in setting up secure, verifiable, scalable and traceable systems. Blockchain manages the audit and contract management processes with ease. Virtual water currency is critical to advocate sustainability. The objective of this paper is to establish the key linkages between virtual water and usage of blockchain. A systematic literature survey was conducted on 16 journal repositories (153 journal papers) of IWA Publishing to establish virtual water linkages and five journal databases (IEEE Xplore, Sciencedirect, ACM Digital Library, Springer Link and Wiley Online Library covering 5026 journal papers) for blockchain and water management linkages. This study proposes to introduce virtual water currency and set up an International Virtual Water Trading System using blockchain. The proposed platform seamlessly integrates the quality, cost and sustainability of industrial products and their sub-components.
Whereas the use of distributed ledger technologies has previously been limited to cryptocurrencies, other sectorsâsuch as healthcare, supply chain, and financeâcan now benefit from them because of bitcoin scripts and smart contracts. However, these applications rely on oracles to fetch data from the real world, which cannot reproduce the trustless environment provided by blockchain networks. Despite their crucial role, academic research on blockchain oracles is still in its infancy, with few contributions and a heterogeneous approach. This study undertakes a bibliometric analysis by highlighting institutions and authors that are actively contributing to the oracle literature. Investigating blockchain oracle research state of the art, research themes, research directions, and converging studies will also be highlighted to discuss, on the one hand, current advancements in the field and, on the other hand, areas that require more investigation. The results also show that although worldwide collaboration is still lacking, various authors and institutions have been working in similar directions.
As cryptocurrency markets such as Bitcoin and Ethereum increase worldwide, cryptocurrency exchanges, service providers, and governments need to monitor and analyze cryptocurrency transactions. There are a few studies to analyze Bitcoin data in open-source projects. However, under massive data volume in the blockchain structure, we still meet the challenges of data processing or analysis speed and efficient data structure. For this purpose, this paper introduces BitSQL, which minimizes significantly database construction time and analysis performance using well-known RDBMS. BitSQL maintains the compact database due to a data mapping technique and a transaction split technique for Bitcoin analysis and it achieves the fast response time due to the index of the database. In addition, as BitSQL is based on the well-known RDBMS, we can easily support web integration or visualization tools. Through experiments, we demonstrate that the data mapping technique can decrease the database size by 95% compared to the simple Bitcoin data storing method. We also show that BitSQL can perform easily analysis regarding balances, transactions, and Bitcoin address clustering heuristics through SQL query statements. This study contributes to the scalable Bitcoin analysis software construction method for analyzing and tracking cryptocurrency blockchains and creating services.
Deep Learning (DL) models are increasingly used for financial forecasting problems, such as price or trend prediction of a financial asset. However, most methods either rely solely on price information or require difficult to implement data harvesting pipelines, e.g., from social media, to deploy them. The main contribution of this paper is a method that exploits sentiment information as a source of additional supervision during the training process, allowing for improving the profitability of the developed strategies compared to baseline agents, while also allowing for operating the agent under partial observability, i.e., without requiring sentiment information as input during inference. As demonstrated in the conducted experiments on the Bitcoin-USD currency pair, this approach can indeed lead to significant improvements in the performance of DL agents, as well as help reduce the overfitting phenomena that often occur when training such agents.
Single point of failure and node attack tend to cause instability in the centralized Internet of Things (IoT). Combined with blockchain technology, the deficiency of traditional IoT architecture can be effectively alleviated. However, the existing blockchain consensus mechanism still has the problems of forks and wasting of computing power. Therefore, this paper proposes a new framework based on a two-stage credit calculation to handle these problems. Notably, the nodes are selected through the model, and these nodes will compete on the chain according to the behavior of participating in the creation of the block. Then, a comparative simulation with the existing consensus mechanism proof of work (PoW) is presented. The results show that the proposed framework can quickly eliminate malicious nodes, maintain the overall security of the blockchain and reduce consensus delay.
Tamima Elbashbishy, Gasser G. Ali, Islam H. El-adaway
Blockchain represents an evolving technology for distributed and secure recording and sharing of information. Meanwhile, blockchain has thrived in banking, finance, and supply chain; its usage within the construction industry is still in its infancy. To this end, the existing literature falls short in providing comprehensive quantitative understanding, within a systems-based analytic context, of the factors affecting blockchain utilization in construction applications. This paper fills this knowledge gap. The authors: (1) conducted an extensive literature review on blockchain implementation in the construction domain; (2) identified a list of 41 factors affecting blockchain implementation in construction projects categorized in four categories: challenges, needs, requirements, and capabilities; (3) utilized a social network analysis (SNA) approach on a database of 111 publications to quantitatively analyze the literature as related to the aforementioned factors; and (4) performed clustering analysis on the SNA graphs to determine the combinations of factors that are most likely co-occurring in research publications. SNA results indicate that while the most investigated factor was âincreased trust and transparency between project partiesâ, the least studied factors included: âcash upfront funding systemâ, âchange payment processes and proceduresâ, âsmart contracts design errorsâ, âcryptocurrency fluctuationsâ, âlack of sufficiently skilled personnelâ, and âincreased awareness and capabilities of personnelâ. Also, clustering outcomes highlight that some combinations of factors are not well-represented in current scholarly efforts. Such imbalance and consequent knowledge gaps may contribute to the actual implementation rate of blockchain in construction applications. Ultimately, this paper provides a roadmap for potential future directions of blockchain construction-related research.
Enrico Tedeschi, Tor-Arne S. Nordmo, Dag Johansen, HĂĽvard D. Johansen
The transaction-rate bottleneck built into popular proof-of-work (PoW)-based cryptocurrencies, like Bitcoin and Ethereum, leads to fee markets where transactions are included according to a first-price auction for block space. Many attempts have been made to adjust and predict the fee volatility, but even well-formed transactions sometimes experience unexpected delays and evictions unless a substantial fee is offered. In this article, we propose a novel transaction inclusion model that describes the mechanisms and patterns governing miners decisions to include individual transactions in the Bitcoin system. Using this model we devise a Machine Learning (ML) approach to predict transaction inclusion. We evaluate our predictions method using historical observations of the Bitcoin network from a five month period that includes more than 30 million transactions and 120 million entries. We find that our Machine Learning (ML) model can predict fee volatility with an accuracy of up to 91%. Our findings enable Bitcoin users to improve their fee expenses and the approval time for their transactions.
Song Tang, Zhiqiang Wang, Jian Jiang, Suli Ge ¡ 5 authors
With the continuous development of blockchain technology, the application scenarios of alliance blockchain are also increasing. The consensus algorithm can achieve distributed consensus among nodes in the network. At present, the practical byzantine fault tolerance algorithm (PBFT) consensus algorithm commonly used in alliance blockchain requires all nodes in the network to participate in the consensus process. Experiments show that when the number of consensus nodes in the system exceeds 100, the bandwidth consumption and consensus delay will greatly increase, resulting in the inability of PBFT to be applied. In scenes with many nodes. How to improve the performance of alliance blockchains safely and efficiently has become an urgent problem to be solved at present. For the PBFT commonly used in alliance blockchains, there are some problems, such as large communication overhead, simple selection of master nodes, and inability to expand and exit nodes dynamically in the network. This paper proposes an improved algorithm tPBFT (trust-based practical Byzantine algorithm), which is suitable for high-frequency trading scenarios of consortium chains. By introducing a trust equity scoring mechanism between nodes in the network, the list of consensus nodes can be dynamically adjusted. tPBFT simplifies the pre-prepare stage of the PBFT consensus process, and realizes the verification of the hash transaction list in the reply stage, thereby reducing the interaction overhead between network nodes. Theoretical analysis and experiments show that when the number of nodes in the network is greater than 30, with the further increase of the number of nodes, the improved tPBFT algorithm has a relatively large performance in terms of node communication overhead, consensus efficiency and scalability outperforms the PBFT algorithm.
Machine learning (ML) has been pervasively researched nowadays and it has been applied in many aspects of real life. Nevertheless, issues of model and data still accompany the development of ML. For instance, training of traditional ML models is limited to the access of data sets, which are generally proprietary; published ML models may soon be out of date without an update of new data and continuous training; malicious data contributors may upload wrongly labeled data that leads to undesirable training results; and the abuse of private data and data leakage also exit. With the utilization of blockchain, an emerging and swiftly developing technology, these problems can be efficiently solved. In this paper, we survey the convergence of collaborative ML and blockchain. Different ways of the combination of these two technologies are investigated and their fields of application are examined. Discussion on the limitations of current research and their future directions are also included.
Activation functions are critical components of neural networks, helping the model learn highly-intricate dependencies, trends, and patterns. Non-linear activation functions allow the model to behave as a functional approximator, learning complex decision boundaries and multi-dimensional patterns in the data. Activation functions can be combined with one another to learn better representations with the objective of improving gradient flow, performance metrics reducing training time and computational cost. Recent work on oscillatory activation functions\cite{noel2021growing}\cite{noel2021biologically} showcased their ability to perform competitively on image classification tasks using a compact architecture. Our work proposes the utilization of these oscillatory activation functions for predicting the volume-weighted average of Bitcoin on the G-Research Cryptocurrency Dataset. We utilize a popular LSTM architecture for this task achieving competitive results when compared to popular activation functions formally used.