In this paper, we evolve heuristics to guide staged deepening search for the hard game of FreeCell, obtaining top-notch solvers for this human-challenging puzzle. We first devise several novel heuristic measures using minimal domain knowledge and then use them as building blocks in two evolutionary setups involving a standard genetic algorithm and policy-based, genetic programming. Our evolved solvers outperform the best FreeCell solver to date by three distinct measures: 1) number of search nodes is reduced by over 78%; 2) time to solution is reduced by over 94%; and 3) average solution length is reduced by over 30%. Our top solver is the best published FreeCell player to date, solving 99.65% of the standard Microsoft 32 K problem set. Moreover, it is able to convincingly beat high-ranking human players.

Diffuse optical tomography (DOT) is a non-invasive method for detecting breast cancer; however, it struggles to produce high-quality images due to the complexity of scattered light and the limitations of traditional image reconstruction algorithms. These algorithms can be affected by boundary conditions and have a low imaging accuracy, a shallow imaging depth, a long computation time, and a high signal-to-noise ratio. However, machine learning can potentially improve the performance of DOT by being better equipped to solve inverse problems, perform regression, classify medical images, and reconstruct biomedical images. In this study, we utilized a machine learning model called “XGBoost” to detect tumors in inhomogeneous breasts and applied a post-processing technique based on genetic programming to improve accuracy. The proposed algorithm was tested using simulated DOT measurements from complex inhomogeneous breasts and evaluated using the cosine similarity metrics and root mean square error loss. The results showed that the use of XGBoost and genetic programming in DOT could lead to more accurate and non-invasive detection of tumors in inhomogeneous breasts compared to traditional methods, with the reconstructed breasts having an average cosine similarity of more than 0.97 ± 0.07 and average root mean square error of around 0.1270 ± 0.0031 compared to the ground truth.

Diffuse optical imaging (DOI) uses scattered light to non-invasively sense and image highly diffuse media, including biological tissues such as the breast and brain. Despite its clinical potential, widespread adoption remains limited because physical constraints, limited available datasets, and conventional reconstruction algorithms struggle with the strongly nonlinear, ill-posed inverse problem posed by multiple photon scattering. We introduce Diffuse optical Imaging using Genetic Programming (DI-GP), a physics-guided and fully interpretable genetic programming framework for DOI. Grounded in the diffusion equation, DI-GP evolves closed-form symbolic mappings that enable fast and accurate 2-D reconstructions in strongly scattering media. Unlike deep neural networks, Genetic Programming (GP) naturally produces symbolic expressions, explicit rules, and transparent computational pipelines—an increasingly important capability as regulatory and high-stakes domains (e.g., FDA/EMA, medical imaging regulation) demand explainable and auditable AI systems, and where training data are often scarce. DI-GP delivers substantially faster inference and improved qualitative and quantitative reconstruction performance compared to analytical baselines. We validate the approach in both simulations and tabletop experiments, recovering targets without prior knowledge of shape or location at depths exceeding ~25 transport mean-free paths. Additional experiments demonstrate centimeter-scale imaging in tissue-like media, highlighting the promise of DI-GP for non-invasive deep-tissue imaging and its potential as a foundation for practical DOI systems.