13 problems

Let an active sequential hypothesis testing (ASHT) problem satisfy the weaker Assumption (IIa), and let Chernoff's Procedure A denote the policy used for sequential experimental de…

A CuBAS-selected warm-up set is a subset of the available training data, chosen using the CuBAS framework's curvature-based sampling procedure. Let the warm-up set comprise $5\%$--…

Let $I(n)$ be the truncated JEPIG term obtained from the first $n$ evaluation labels, and let $I$ denote its limit as $n\to\infty$, when this limit exists. The evaluation data dist…

Weak-instrument conjecture. Active learning approaches could help address the problem of weak instruments by selecting values of $Z$ that induce maximal variation in $X$.

Sufficiency conjecture. The upper bound of Lemma should suffice to invoke the cited matrix-completion result, although with larger residual terms and a lower probability of success…

Clique algorithm optimality conjecture. For an $n$-set with a random partition with probabilities $p_1,p_2,,,,,,,,,,$, the clique algorithm has minimal average complexity among all…

The courses considered are mathematics courses offered in English as a medium of instruction in a Confucian heritage culture (CHC-EMI) environment. Interactive engagement is a peda…

Let a stream algorithm emulate a general utility-based pool algorithm, and suppose it is allowed a constant-factor increase in the number of element selections. Constant-factor imp…

The paper studies active learning under the Berkson errors-in-variables model, in which an observed feature $W$ is perturbed to $X=W+\delta$ before the oracle returns a label depen…

Let $cmathbb{C}$ be a binary hypothesis class with VC dimension $d$ and star number $\mathfrak{s}$. For $a dots$ and $mathcal{P}$ satisfying the Bernstein condition with parameters…

Activized empirical risk minimization conjecture. For any VC class, there exists an active meta-algorithm $\mathcal A_a$ and an empirical risk minimization algorithm $\mathcal A_p$…

Rate conjecture. The correct rate of convergence for the excess risk is

Consider a finite hypothesis class ${\cal H}$ with $N=|{\cal H}|$, and suppose a noiseless generalized binary search procedure uses $n_0$ queries. In the noisy setting, each query…