This post is about a technology which could develop in the future.<\/p>\n
Right now, a new drug might be tested by finding patients with some diagnosis and giving or not giving them a drug according to a secret randomization. The outcome is observed, and if the average outcome for those treated is measurably better than the average outcome for those not treated, the drug might become a standard treatment.<\/p>\n
Generalizing this, a filter F<\/em> sorts people into two groups: those for treatment A<\/em> and those not for treatment B<\/em> based upon observations x<\/em>. To measure the outcome, you randomize between treatment and nontreatment of group A<\/em> and measure the relative performance of the treatment.<\/p>\n A problem often arises: in many cases the treated group does not do better than the nontreated group. A basic question is: does this mean the treatment is bad? With respect to the filter F<\/em> it may mean that, but with respect to another filter F’<\/em>, the treatment might be very effective. For example, a drug might work great for people which have one blood type, but not so well for others.<\/p>\n Finding F’<\/em> is a situation where machine learning can help. The setting is essentially isomorphic to this one<\/a>. The basic import is that we can learn a rule F’<\/em> for filters which are more strict than the original F<\/em>. This can be done on past recorded data<\/em>, and if done properly we can even statistically prove that F’<\/em> works, without<\/em> another randomized trial. All of the technology exists to do this now—the rest is a matter of education, inertia, and desire.<\/p>\n Here’s what this future might look like:<\/p>\n This post is about a technology which could develop in the future. Right now, a new drug might be tested by finding patients with some diagnosis and giving or not giving them a drug according to a secret randomization. The outcome is observed, and if the average outcome for those treated is measurably better than … <\/p>\n\n