Publisher Summary As the Induced electrical polarization (IP) technique provides very useful information about the properties of geologic materials and the time-domain electromagnetic (TEM) technique provides better field data acquisition, it is important to study the extraction of the IP information from the TEM data. The knowledge of the electrical behavior of the heterogeneous materials helps greatly in improving the accuracy of the interpretation. The TEM response measured using a coincident loop system above a dispersive conductive earth can show evidence of IP effects that manifest themselves as a negative response (NR) phenomenon where the transient voltage response undergoes a rapid decay, which is followed by a polarity reversal. The negative response is regarded as noise by many practicing geophysicists and eliminated from their field data because there exists no convenient way of inverting the IP effect. Much research has been done on the inversion of EM measurements above polarizable ground. A number of methods have been used for determining the chargeability from the time-domain or frequency-domain IP data using an electrode system. Inversion methods have been developed for estimating the Cole-Cole parameters from time-domain IP data. This chapter presents a neural network approach for finding the electrical properties of half-space and layered polarizable conducting targets using transient electromagnetic coincident or central loop systems to predict the Cole-Cole parameters.