本次英国代写是一个C++字符串清理相关毕设算法程序

Sequences are often analyzed in applications including location-based service provision, product recommendation, and DNA sequence analysis but this may lead to privacy breaches.

The focus of all 5 sub-topics below is the development of algorithms and their experimental evaluation. Strong knowledge of data structures and algorithms is needed, as well as strong programming skills (preferably in C++).

- The task is to protect the presence or the absence of all q-grams (substrings of q letters) in a given sequence by applying differential privacy: https://www.cis.upenn.edu/~aaroth/Papers/privacybook.pdf (see Sections 1,2, and 3).
- The task is to identify how to replace #s introduced by the TFS algorithm in ttps://www.ecmlpkdd2019.org/downloads/paper/73.pdf while minimizing tau-ghosts. See https://hal.archives-ouvertes.fr/hal-03070560/document
- The task is to try to “break” the output of MCSR in https://www.ecmlpkdd2019.org/downloads/paper/73.pdf by guessing where the #s might have been added and/or likely letters that #s in X corresponded to. See Section 6.3 of https://dl.acm.org/doi/pdf/10.1145/3418683
- Consider a set of sequences X={X1,…,Xn} and a set of subsequences S that model confidential information. The Minimum Utility Loss Generalization problem seeks to transform X with minimum utility loss so that a notion of privacy, based on mutual information, is satisfied. The problem has been introduced in https://dl.acm.org/doi/pdf/10.1145/2835776.2835828, where also heuristics have been proposed. The task is to implement, evaluate, and potentially adapt the heuristics to different problems.
- The top-k selection problem requires selecting the “best” k elements of a given set of elements U according to a quality function. Assume that each element of U is contributed by a different individual. Differential privacy can guarantee that the output k elements will not differ significantly (in a probabilistic sense) based on the input of any individual. The task is to implement and evaluate mechanisms that solve the top-k selection problem under differential privacy.

See https://proceedings.neurips.cc/paper/2020/file/01e00f2f4bfcbb7505cb641066f2859b-Paper.pdf *** Discussion on topics