Ildar Khabibrakhmanov

Analytics-driven solutions for customer targeting and sales-force allocation

Sales professionals need to identify new sales prospects, and sales executives need to deploy the sales force against the sales accounts with the best potential for future revenue. We describe two analytics-based solutions developed within IBM to address these related issues. The Web-based tool OnTARGET provides a set of analytical models to identify new sales opportunities at existing client accounts and noncustomer companies. The models estimate the probability of purchase at the product-brand level. They use training examples drawn from historical transactions and extract explanatory features from transactional data joined with company firmographic data (e.g., revenue and number of employees). The second initiative, the Market Alignment Program, supports sales-force allocation based on field-validated analytical estimates of future revenue opportunity in each operational market segment. Revenue opportunity estimates are generated by defining the opportunity as a high percentile of a conditional distribution of the customers spending, that is, what we could realistically hope to sell to this customer. We describe the development of both sets of analytical models, the underlying data models, and the Web sites used to deliver the overall solution. We conclude with a discussion of the business impact of both initiatives.

Machine learning based multi-physical-model blending for enhancing renewable energy forecast - improvement via situation dependent error correction

With increasing penetration of solar and wind energy to the total energy supply mix, the pressing need for accurate energy forecasting has become well-recognized. Here we report the development of a machine-learning based model blending approach for statistically combining multiple meteorological models for improving the accuracy of solar/wind power forecast. Importantly, we demonstrate that in addition to parameters to be predicted (such as solar irradiance and power), including additional atmospheric state parameters which collectively define weather situations as machine learning input provides further enhanced accuracy for the blended result. Functional analysis of variance shows that the error of individual model has substantial dependence on the weather situation. The machine-learning approach effectively reduces such situation dependent error thus produces more accurate results compared to conventional multi-model ensemble approaches based on simplistic equally or unequally weighted model averaging. Validation over an extended period of time results show over 30% improvement in solar irradiance/power forecast accuracy compared to forecasts based on the best individual model.

Analytical model of the heliopause

Recent multidimensional numerical simulations [Pauls et al., 1995] of the solar wind interaction with the local interstellar medium have produced a self-consistent description of the interstellar neutral hydrogen at the heliopause. In this report, we develop a simple one-dimensional model of the solar wind interaction with interstellar hydrogen which is capable of describing the basic features of the interaction along the solar wind stagnation line that were obtained in the numerical simulations. It is shown that pickup ions in the subsonic solar wind can stagnate the plasma flow completely at a distance of 40 AU from the termination shock thus defining the position of the contact discontinuity relative to the termination shock. The plasma number density increases considerably but remains finite on the solar wind side of the contact discontinuity, for a finite velocity of the neutral hydrogen. For the interstellar wind, deceleration due to the interaction with solar hydrogen is found to be important, even though the solar hydrogen density flux is relatively small. The characteristic scale of this interaction is defined by the depth of penetration of the solar hydrogen atoms into the interstellar wind. We find that for observed solar wind and interstellar medium parameters, the deceleration is sufficient to stagnate the interstellar wind at the contact discontinuity.

Operations Research Improves Sales Force Productivity at IBM

In 2004, IBM introduced a set of broad operations research-based initiatives designed to improve the efficiency and productivity of its global sales force. The first solution, OnTARGET, provides a set of analytical models designed to identify new sales opportunities at existing IBM accounts and at noncustomer companies. The second solution, the Market Alignment Program (MAP), optimally allocates sales resources based on field-validated analytical estimates of future revenue opportunities in operational market segments. IBM Research developed the operations research models and initial internal websites for both solutions. The IBM Software Group initially implemented OnTARGET, which was subsequently made available to over 13,000 sales representatives across IBM sales organizations worldwide. The IBM Sales and Distribution organization deployed MAP as an integral part of its sales model to better align sales resources with the best market opportunities. We describe the development of both analytical models, and the underlying data models and websites used to deliver the solutions. We conclude with a discussion of the business impact, which we estimate as hundreds of millions of dollars annually for the combined initiatives.

Kinetic description of ionospheric outflows based on the exact form of Fokker‐Planck collision operator: Electrons

[1] We present the results of a finite difference implementation of the kinetic Fokker-Planck model with an exact form of the nonlinear collisional operator. The model is time dependent and three-dimensional; one spatial dimension and two in velocity space. The spatial dimension is aligned with the local magnetic field, and the velocity space is defined by the magnitude of the velocity and the cosine of pitch angle. An important new feature of model, the concept of integration along the particle trajectories, is discussed in detail. Integration along the trajectories combined with the operator time splitting technique results in a solution scheme which accurately accounts for both the fast convection of the particles along the magnetic field lines and relatively slow collisional process. We present several tests of the model’s performance and also discuss simulation results of the evolution of the plasma distribution for realistic conditions in Earth’s plasmasphere under different scenarios.

On the usefulness of solar energy forecasting in the presence of asymmetric costs of errors

Because of the weather-associated variability of renewable energy generation, forecasting is an inherent component of an overall solution to reduce the grid integration cost of renewable energy. Accuracy of a forecast is characterized typically by metrics such as root mean square error (RMSE) or mean absolute error (MAE). Such metrics, however, may not comprehensively capture the usefulness of a forecast. Use cases of forecasts are usually complex and are connected to how energy producers, balancers, or traders may apply the forecast to minimize some cost functions in order to improve performance. Often a cost function is asymmetric, unlike RMSE or MAE. Here, we treat complex cost functions as asymmetric perturbations to the symmetric RMSE metric and ask how a forecast can be statistically corrected to minimize the cost function. The analysis leads to an analytical expression for one aspect of a forecasts usefulness, which characterizes the capability of a user to benefit from the knowledge of the asymmetry of the cost function. As a case study, we present a comparison of solar forecasts derived from a number of numerical weather predictions at a test site in Rutland, Vermont.

On the theory of Alfvén waves in the solar wind

Parallel propagating finite-amplitude Alfven waves in a nonhomogeneous medium are considered. The analysis is based on an ideal magnetohydrodynamic (MHD) description with an arbitrary geometry for the background magnetic field. Examination of the characteristic equations of the system reveals, as expected, six waves: forward and backward propagating Alfven waves and fast/slow magnetosonic waves. However, although these waves are well classified in a homogeneous medium (in particular, the pure Alfven wave is strictly incompressible and circularly polarized), such a simple classification with regard to compressibility and polarization is found to be impossible in a nonhomogeneous medium. We derive exact forms for the wave action invariant equation and the amplitude conservation law for Alfven waves and find these equations to be inconsistent, reaffirming that the pure incompressible Alfven wave does not exist in a nonhomogeneous medium. Construction of a consistent Alfven wave solution is outlined that incorporates predominantly large-amplitude incompressible Alfven waves with small-amplitude compressible fluctuations. Such a solution corresponds to strong fluctuations in the magnetic field vector and weak fluctuations in the magnetic field strength, as typically observed in the solar wind. In order to counteract the tendency of compressible fluctuations to grow without bound in the ideal MHD model, we include a Landau damping term obtained from kinetic theory.