Mixed-Integer Linear Programming (MILP) 
This paper proposes an optimization technique to help utility operators to recuperate energy distribution methods after giant outages. Particularly, a novel mixed-integer linear programming (MILP) mannequin is developed for co-optimizing crews, sources, and community operations. The MILP mannequin coordinates the injury isolation, community reconfiguration, distributed generator re-dispatch, and crew/useful resource logistics. We think about two several types of crews, specifically, line crews for injury restore and tree crews for impediment removing. We additionally mannequin the restore useful resource logistic constraints. Moreover, a brand new algorithm is developed for fixing the distribution system restore and restoration downside (DSRRP). The algorithm begins by fixing DSRRP utilizing an assignment-based technique, then a neighborhood search technique is designed to iteratively enhance the answer. The proposed technique is validated on the modified IEEE 123-bus distribution check system. …
Bayes via Goodness of fit
The 2 key points of contemporary Bayesian statistics are: (i) establishing principled method for distilling statistical prior that’s according to the given knowledge from an preliminary plausible scientific prior; and (ii) growth of a Bayes-frequentist consolidated knowledge evaluation workflow that’s simpler than both of the 2 individually. On this paper, we suggest the thought of ‘Bayes through goodness of match’ as a framework for exploring these elementary questions, in a means that’s normal sufficient to embrace nearly the entire acquainted chance fashions. A number of illustrative examples present the good thing about this new standpoint as a sensible knowledge evaluation software. Relationship with different Bayesian cultures can be mentioned. …
Quantile Double Autoregression
Many monetary time collection have various buildings at completely different quantile ranges, and in addition exhibit the phenomenon of conditional heteroscedasticity on the identical time. In the intervening time, it’s nonetheless lack of a time collection mannequin to accommodate each of the above options concurrently. This paper fills the hole by proposing a novel conditional heteroscedastic mannequin, which is named the quantile double autoregression. The strict stationarity of the brand new mannequin is derived, and a self-weighted conditional quantile estimation is recommended. Two promising properties of the unique double autoregressive mannequin are proven to be preserved. Based mostly on the quantile autocorrelation perform and self-weighting idea, two portmanteau assessments are constructed, they usually can be utilized in conjunction to verify the adequacy of fitted conditional quantiles. The finite-sample efficiency of the proposed inference instruments is examined by simulation research, and the need of the brand new mannequin is additional demonstrated by analyzing the S&P500 Index. …
Destruction Rate
It’s troublesome to detect and take away secret photographs which might be hidden in pure photographs utilizing deep-learning algorithms. Our method is the primary work to successfully disable covert communications and transactions that use deep-learning steganography. We deal with the issue by exploiting refined pixel distributions and edge areas of photographs utilizing a deep neural community. Based mostly on the given info, we adaptively take away secret info on the pixel stage. We additionally introduce a brand new quantitative metric referred to as destruction price because the decoding technique of deep-learning steganography is approximate (lossy), which is completely different from standard steganography. We consider our method utilizing three public benchmarks compared with standard steganalysis strategies and present that the decoding price improves by 10 ~ 20%. …
