Advancements in Decision-Making Algorithms and Blockchain Security

The recent publications in the field of computational decision-making and blockchain technology reveal a significant push towards enhancing the efficiency, security, and scalability of systems through innovative algorithmic approaches and formal verification methods. A notable trend is the development of concise and efficient representations of complex decision-making policies, such as small decision trees for Markov Decision Processes (MDPs), which aim to optimize computational resources while maintaining or improving decision accuracy. Additionally, there is a growing emphasis on the formal verification of algorithms and systems, particularly in the context of MDPs and blockchain technologies, to ensure their correctness and reliability. This includes the use of SMT-based approaches and the integration of formal verification tools like Isabelle/HOL to provide rigorous proofs of system properties.

In the realm of blockchain and decentralized systems, research is increasingly focused on addressing security vulnerabilities, such as network packet-based cheats and selfish mining strategies, through novel cryptographic protocols and mechanism designs. The exploration of decentralized federated learning frameworks for detecting malicious activities, such as financial bots, highlights the field's move towards leveraging distributed ledger technologies for enhanced security and privacy. Furthermore, the development of comprehensive datasets and benchmark platforms for Ethereum and other blockchain ecosystems underscores the importance of data-driven approaches in advancing blockchain analytics and decentralized finance research.

Noteworthy Papers

  • Small Decision Trees for MDPs with Deductive Synthesis: Introduces an SMT-based approach and an abstraction-refinement loop for encoding optimal policies as small decision trees, significantly reducing the size of policy representations.
  • A Formally Verified IEEE 754 Floating-Point Implementation of Interval Iteration for MDPs: Presents a formally verified implementation of interval iteration for MDPs, ensuring the correctness of computations involving floating-point arithmetic.
  • BotDetect: A Decentralized Federated Learning Framework for Detecting Financial Bots on the EVM Blockchains: Proposes a decentralized federated learning approach for detecting financial bots, enhancing security and privacy in blockchain ecosystems.
  • Mechanism Design for Blockchain Order Books against Selfish Miners: Offers an analytical study and a mechanism to mitigate the social welfare loss caused by selfish miners in blockchain-based order book systems.

Sources

Small Decision Trees for MDPs with Deductive Synthesis

A Formally Verified IEEE 754 Floating-Point Implementation of Interval Iteration for MDPs

Visualizing Game-Based Certificates for Hyperproperty Verification

A Framework for Mining Collectively-Behaving Bots in MMORPGs

Addressing Network Packet-based Cheats in Multiplayer Games: A Secret Sharing Approach

Automated Selfish Mining Analysis for DAG-based PoW Consensus Protocols

Blockchain-assisted Demonstration Cloning for Multi-Agent Deep Reinforcement Learning

An Analysis of the Correctness and Computational Complexity of Path Planning in Payment Channel Networks

Fixed Point Certificates for Reachability and Expected Rewards in MDPs

Characterizing Transfer Graphs of Suspicious ERC-20 Tokens

Multi-source Multi-level Multi-token Ethereum Dataset and Benchmark Platform

BRC20 Snipping Attack

BotDetect: A Decentralized Federated Learning Framework for Detecting Financial Bots on the EVM Blockchains

Mechanism Design for Blockchain Order Books against Selfish Miners

Chasing price drains liquidity

Built with on top of