Near-optimal linear predictive clustering in non-separable spaces is achieved through MIP complexity reductions with provable bounds and QPBO approximations that outperform greedy methods in regression error and scalability.
2024)—against our proposed scalable QBPO method that uses Gurobi’s MIP solver, encoding identical problem instances in the OPB format under uniform conditions
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Near-optimal Linear Predictive Clustering in Non-separable Spaces via MIP and QPBO Reductions
Near-optimal linear predictive clustering in non-separable spaces is achieved through MIP complexity reductions with provable bounds and QPBO approximations that outperform greedy methods in regression error and scalability.