Privacy preservation is important. Prescriptive analytics is a method to extract corrective actions to avoid undesirable outcomes. We propose a privacy preserving prescriptive analytics algorithm to protect the data used during the construction of the prescriptive analytics algorithm. We use differential privacy mechanism to achieve strong privacy guarantee. Differential privacy mechanism requires computation of sensitivity: maximum change in the output between two training datasets, which is differed by only one instance. The main challenge we addressed is the computation of sensitivity of the prescription vector. In absence of any analytical form, we construct a nested global optimization problem to compute the sensitivity. We solve the optimization problem using constrained Bayesian optimization, as the nested structure makes the objective function expensive. We demonstrate our algorithm on two real world datasets and observe that the prescription vectors remains useful even after making them private.

Head yaw angle is an important indicator of a driver's distraction. Forward collision warning systems incorporate head yaw angle estimators for detecting driver's head-off-road glances. This paper presents a generic geometric model for head yaw estimation. The generic model is customized into 12 different models. The performance of the cylindrical and ellipsoidal models is improved by introducing nose projection and truncation. The yaw estimation problem is formulated using a single variable. This formulation provides better visualization of the cylindrical and ellipsoidal models. It is shown that the ellipsoidal model in literature is indeed a cylindrical model with nose projection. We have also proved that, under the ellipsoidal/cylindrical framework, the estimated yaw angle is independent of shifting the center of rotation along depth. Four previous works are shown as the customization of the proposed generic model. The performance of the proposed models are evaluated using four standard head pose datasets and an ADAS dataset. The proposed nose-projected truncated ellipsoidal model outperforms the state-of-the-art geometric models. Further, we have studied the impact of model parameters and inaccurate facial feature localization through simulation experiments.

Accurate head yaw estimation is necessary for detecting driver inattention in forward collision warning systems. In this paper, we propose three geometric models under the ellipsoidal framework for accurate head yaw estimation. We present theoretical analysis of the cylindrical and ellipsoidal face models used for yaw angle estimation of head rotation. The relationship between cylindrical, ellipsoidal, and proposed models is derived. We provide error functions for all models. Furthermore, for each model, over/under estimation of angle, zero crossings of error, bounds on yaw angle estimate, and bounds on error are presented. Experimental results of the proposed models on four standard head pose data sets yielded a mean absolute error between 4° and 8° demonstrating the efficacy of the proposed models over the state-of-the-art methods.

Most of the design techniques of T-S fuzzy controllers assumes that there exists an approximate T-S model of the system with fixed antecedent parts & rules and uses techniques like GA, LMI, etc for the optimal design of the gain values. This paper presents a novel integrated approach for the design and stability analysis of T-S fuzzy controllers using GA with variable length chromosomes (VLCs) and LMI. This approach helps to find out the optimal parameters of the antecedent parts of the rules along with rule optimization and also to optimize the consequent parts.

In this paper, a shape adaptive wavelet transform coding scheme for arbitrarily shaped visual objects which is also scalable and progressive is proposed. The proposed scheme uses the shape adaptive wavelet transform (SA-DWT) of extracted objects for generating the wavelet coefficients and the wavelet difference method (SWDR) for progressive scalable coding and decoding. The method is developed for the limited bandwidth network where the image quality and data compression are most important. The experiments are performed on the MRI medical images, and video sequence images. The simulation results show that the proposed scheme is slightly better at all bit rates, in PSNR values compared to the other scalable coding schemes like SPIHT etc. Thus, the proposed coding scheme gives a convenient and better object coding scheme with applications in the multimedia image compression and transmission scenario.

A fuzzy approach was proposed in the above paper for the solution of strategic games and some useful conclusions were presented on the selection of optimal strategies in the prisoner's dilemma game. In this paper, a general result is presented for the case in which a prisoner is noncooperative and selfish, which makes the evaluation suggested in the above paper redundant in this case.

In this paper, optimal fuzzy observer is developed for discrete time nonlinear stochastic system based on Takagi-Sugeno (TS) fuzzy model. We design a optimal fuzzy observer using the mathematical duality and by extending the nonlinear Bayesian estimation using Gaussian sum approach to the TS fuzzy model.

This paper presents the exact fuzzy modeling and optimal control of a class of second order nonlinear dynamic systems. Input saturation, output slew rate limit and nonlinear stiffness are considered. Conventionally, the TSK fuzzy modeling blends local linear models to represent a nonlinear system, which in general does not exactly represent the nonlinear system under consideration. Here, instead of local linear models, a set of 'boundary linear models' (BLM) and associated membership-functions are so chosen that the fuzzy blending of these models result in an exact representation of the overall nonlinear system. An optimal fuzzy controller is designed based on this exact fuzzy model and the results are compared with a conventional design.

This paper presents a novel method of Exact Fuzzy Modeling and Optimal control of a Launch Vehicle in the atmospheric phase. Time varying plant is considered with control power plant nonlinearity. Plant modeling, controller design and analysis are carried out for Rigid Body with Actuator dynamics. The results are compared with a classical design. Conventionally, the TSK fuzzy modeling blends local linear models to represent a nonlinear system, which in general does not exactly represent the nonlinear system under consideration. Here, instead of local linear models, a set of 'boundary linear models' and associated membership-functions are so chosen that the fuzzy blending of these models result in an exact representation of the overall nonlinear time varying system. Optimal control laws are designed for each of these fuzzy subsystems, and overall control is again the fuzzy blending of these individual control laws. This procedure results in an exact optimal control solution for the Launch Vehicle control in the atmospheric phase.

This paper presents a fast encoding algorithm for vector quantization that uses edge information contained in an image in addition to the use of average values of the vectors forming the image. The algorithm makes use of these features to identify the visually significant areas of the image and to save computational time while producing good quality pictures at reasonable computing cost. The simulation results compare well with other existing algorithms of the conventional type.

We present a perfect balancing method for a binary search tree. During the updates the algorithm allows the structure to grow gracefully and maintains the optimal shape without degeneration. The algorithm uses swapping as the basic operation. Since the tree produced by the algorithm is optimal it can favourably be compared with that produced by other balancing algorithms. In worst case situation, the algorithm takes O(n) time, n being the total number of nodes in the tree. This is an added significance when it is compared with the static optimal binary search trees.