Sleep quality influences health, cognition, and daily functioning, yet no universally accepted objective measure exists. This thesis introduces a transparent Sleep Quality (SQ) metric derived from polysomnography (PSG) data. First, eleven key PSG features were identified from a comprehensive literature review and scored against age-specific clinical reference ranges (0 = poor, 2 = good). Each feature score was normalized to a 0–1 scale and combined via a linear weighted sum, with weights learned by linear regression on the Sleep Heart Health Study cohort (7,765 adults) to match subjective morning ratings. The final four-feature model (Total Sleep Time (TST), Sleep Efficiency (SE), Arousal Index (AI), Sleep Stage Distribution (SSD)) explained only about 6.7 % of variance in subjective ratings (R2 ≈ 0.067, RMSE ≈ 0.224), indicating limited alignment between objective and perceived SQ. This work provides an interpretable framework for objective sleep assessment, showing how to select metrics, how to give them proper scores, as well as how to find proper weights for an objective SQ metric.