Looping weighted k-NN algorithym for constructing missing feature values for cancer

Healthcare facilities have at their disposal vast amounts of cancer patients’ data. Medical practitioners require more efficient techniques to extract relevant knowledge from this data for accurate decision-making. However the challenge is how to extract and act upon it in a timely manner. If well engineered, the huge data can aid in developing expert systems for decision support that can assist physicians in diagnosing and predicting some debilitating life threatening diseases such as cancer. Expert systems for decision support can reduce the cost, the waiting time, and liberate medical practitioners for more research, as well as reduce errors and mistakes that can be made by humans due to fatigue and tiredness. The process of utilizing health data effectively however, involves many challenges such as the problem of missing feature values, the curse of dimensionality due to a large number of attributes, and how to go about determining the features that can lead to more accurate diagnosis. Effective data mining tools can assist in early detection of diseases such as cancer. In this paper, we bring forth a new approach for constructing missing features values based on iterative nearest neighbours and distance metrics. This approach employs weighted k-nearest neighbour's algorithm. This integrates Euclidean and Minkowski functions as a component toward looping k-NN classifier. Our major inspiration is to propagate the classification accuracy to a certain threshold that is set by either researchers or users.