China
Africa
Explore chapters and articles related to this topic
Logistic Regression: The Workhorse of Response Modeling
Published in Bruce Ratner, Statistical and Machine-Learning Data Mining, 2017
There are interesting bits of information in Table 10.11 that illustrate two useful statistical identities: Exponentiation of both sides of Equation 10.12 produces odds of response equal to 0.1358. Recall, exponentiation is the mathematical operation of raising a quantity to a power. The exponentiation of a logit is the odds. Consequently, the exponentiation of −1.9965 is 0.1358. See Equations 10.13 through 10.15.Exp(Logit(TNX_ADD=1))=Exp(−1.9965)Odds(TNX_ADD=1)=Exp(−1.9965)Odds(TNX_ADD=1)=0.1358The probability of (TNX_ADD = 1), hereafter the probability of RESPONSE, is easily obtained as the ratio of odds divided by 1 + odds. The implication is that the best estimate of RESPONSE—when no information or variables are available—is 11.9%, namely, the average response of the mailing.
A novel appliance-based secure data aggregation scheme for bill generation and demand management in smart grids
Published in Connection Science, 2021
Yihui Dong, Jian Shen, Sai Ji, Rongxin Qi, Shuai Liu
The computational overhead of users and the CC are discussed in this section. Let , , and denote the computational overhead of an exponentiation operation in , an exponentiation operation in G, and a pairing operation, respectively. Since the addition and multiplication operations in is negligible relative to the above operations, we don't take these operations into account. In addition, the computational overhead of an exponentiation operation in a multiplicative group is approximately equal to that of a multiplication operation in an additive group. So, is also utilised to denote a multiplication operation in additive group. Note that n is the number of users in each area, is the number of areas. We use our scheme to do the analysis and compare it with EPPDR (H. Li et al., 2014) and ECBDA (Vahedi et al., 2017).
An efficient attribute-based encryption scheme based on SM9 encryption algorithm for dispatching and control cloud
Published in Connection Science, 2021
Honghan Ji, Hongjie Zhang, Lisong Shao, Debiao He, Min Luo
Table 2 shows the time consumption of setup phase in SM9-ABE takes the least time, making it faster to start up such a system. In key generation phase and encryption phase, we use scalar multiplication in and instead of exponentiation operation, which is of course more efficient. Then it comes to the decryption phase. It seems that Jiang's scheme (Jiang & Susilo, 2018) takes the most time because S is usually larger than . In fact, the latter three schemes all use the Lagrange Interpolation Formula to calculate the secret values. The cases are too complex to calculate the time cost so we ignore them. Therefore, the time cost in decryption phase is nearly equal among the four schemes.
Deniable authenticated encryption for e-mail applications
Published in International Journal of Computers and Applications, 2020
Chunhua Jin, Guanhua Chen, Changhui Yu, Jianyang Zhao
We compare our scheme with other existing DAE schemes [17–19] in terms of computational cost, communication overhead, and security in Table 1. We denote by PM the point multiplication in , ADD the point add in , EXP the exponentiation in , P the pairing computation in and CV the certificate verification (which normally costs about two EXP computation). We note that modular exponentiation operation in finite field is equivalent to point multiplication operation in ECC (i.e. ), and multiplication operation in finite field is equivalent to point addition operation in ECC (i.e. ). The other operations are omitted since they are trivial.