Canadian hi-tech company offers $45,000 for the best algorithm for identification of substances from electromagnetic signatures. —————————————— FIND Technologies Inc. is a Canadian company that owns novel sensor technology for measuring electromagnetic signatures of materials. The sensor is a robust, inexpensive instrument that detects passive electromagnetic emission from all matter. It has biomedical, homeland security, engineering, geological, and other applications. In order to provide real-time, automatic identification of materials, it is necessary to equip the sensor with intelligent software that would take the electromagnetic measurements and recognize various patterns – “fingerprints” –  embedded in the recordings, which uniquely identify substances present in the material under examination. For the sake of highest quality of the algorithm, the company decided to outsource, or better crowd-source, its development to a large worldwide group of specialists, through an online competition launched at TunedIT platform, with the main prize of $45,000. The task is simple: your algorithm has to assign labels of three different substances to 1500 test samples. If at least 95% of labels are correct, $45,000 is yours! FIND already possesses an algorithm for this task, which achieves accuracy of 79%. According to Dr. Frank LaBella PhD, founder of FIND: “In our laboratory we now rely on pattern recognition of images generated from the signals acquired from the sensor.  For example, different patterns are generated from small amounts of chemicals at virtually any distance from the sensor and even when concealed: from small discrete areas of the human body; from body organs in different stages of activity; from abnormal areas of skin; from tiny amounts of cultured cells, normal or diseased; from metal that is under stress; from metal salts mixed in with soil or sand; from aqueous solutions of specific chemicals.” In the struggle to further improve their software, the company realized that their expertise in data mining and advanced intelligent algorithms is too limited to come up with any better solution. Biomedical Commercialization Canada, an investor and strategic partner with FIND, suggested the  concept of crowdsourcing, which appealed to them very much because of easy access to large pool of talented programmers and experienced specialists. They decided to use this approach in order to seek even better algorithm for their product. FIND selected TunedIT, which connects hi-tech companies seeking most efficient algorithms, with a large world-wide community of computer scientists and algorithm designers, to launch an open competition on the first online platform for the development of intelligent algorithms. Launching a contest with TunedIT guarantees that several hundred professionals will approach the task, and each of them will try different methods and ideas, so whatever they manage to find, one can be sure this will be the best solution ever possible. The competition, entitled Materials Identification Based on Measurements of Passively Emitted Electromagnetic Radiation, will last 6 months or until the 95% threshold is achieved. Every programmer, statistician or data scientist can participate. Web page: http://tunedit.org/challenge/material-classification ——— TunedIT is the first online laboratory dedicated to development of intelligent algorithms through crowdsourcing – the easiest and most cost-efficient way of performing Research & Development, in massively parallel manner. TunedIT specializes in the fields of Data Mining, Machine Learning, Computational Intelligence and Statistical Modeling. TunedIT runs online data mining competitions that address selected real-world problems of data analysis. Contests are open to the whole scientific community and attract hundreds of teams every time. TunedIT   maintains close ties with the scientific community, by organizing competitions for major conferences, like the IEEE International Conference on Data Mining, and providing the platform without charge for educational purposes.