E-Voting: Blockchain Based Voting For Secure Vote Casting

A blockchain is decentralized immutable ledger technology maintaining integrity. So to conduct tamperproof election it’s one of the approach towards it. Smart contracts are Self executed code that is written on Ethereum platform in blockchain. An E-voting system should be completely secure and does not allow voting twice that is double spending in blockchain. So it should be completely transparent. In research work electronic voting application is implemented and tested using smart contract on Ethereum platform with the help of metamask wallet. The results of ballots and votes will be stored on Ethereum blockchain with the help of consensus algorithm proof of stake. This consensus is used in validating a transaction with concept of majority approval. Current electronic voting system requires a centralized authority to control the procedure from ballot input to result output and for monitoring of election. While blockchain technology provide decentralized system which is open across connected nodes. Blockchain assets provide increased level of system security from hacking and fraud. Every transaction in blockchain is time-stamped and signed digitally with the help of cryptographic algorithms, and it assigns unique hash value to every block so it can be trace easily. Blockchain technology is one of solutions because it embraces a decentralized system and the entire databases are owned by many users. The blockchain technology also has much vulnerability due to which many attacks like 51% attack, Double Spending attack, DDOS attack, Sybil attack, Eclipse attack and Routing attack can be performed on it.

Sharp extremal bounds on information diffusion capacities of mobile sensor networks

In this paper, we prove that the maximum possible information diffusion capacity of a mobile sensor network with $n$ agents is equal to $1-\frac{n-2}{n^2}$, and the minimum possible capacity is equal to $\frac{2n+2}{3n}$, assuming that each pair of agents communicates exactly once. Our upper bound shows that the collector-distributor construction of (Gu et al, 2018) attains the maximum possible capacity of any mobile sensor network where agents are limited to communicating exactly once. For mobile sensor networks where agents can communicate any number of times and there are a total of $m$ communications, we prove that the maximum possible capacity is equal to $1-\frac{\binom{n-1}{2}}{m n}$ for $m \geq n$ and we also prove that the minimum possible capacity is equal to $\frac{2}{n}$. We also discuss an error in the proof from (Gu et al, 2018) that the capacity in the random combinatorial model approaches $1$ with high probability as the size of the network increases. However, we show that their gossip algorithm proof technique works for a less restricted version of the model.

A Framework for Kernel-Based Multi-Category Classification

A geometric framework for understanding multi-category classification is introduced, through which many existing 'all-together' algorithms can be understood. The structure enables parsimonious optimisation, through a direct extension of the binary methodology. The focus is on Support Vector Classification, with parallels drawn to related methods. The ability of the framework to compare algorithms is illustrated by a brief discussion of Fisher consistency. Its utility in improving understanding of multi-category analysis is demonstrated through a derivation of improved generalisation bounds. It is also described how this architecture provides insights regarding how to further improve on the speed of existing multi-category classification algorithms. An initial example of how this might be achieved is developed in the formulation of a straightforward multi-category Sequential Minimal Optimisation algorithm. Proof-of-concept experimental results have shown that this, combined with the mapping of pairwise results, is comparable with benchmark optimisation speeds.