Model a problem of recommending house to family as an entropy$-$based semi$-$supervised learning. Can the

output be improved with combination of semi$-$supervised learning and genetic algorithm? How? Make

suitable assumptions. Provide pseudo code.