The Muon $g-2$ Experiment operated at Fermi National Accelerator Laboratory (FNAL, or Fermilab) between 2018 and 2023 to produce the world's most precise measurement of the muon's \textit{anomalous magnetic moment}, $a_\mu = \frac{g_\mu - 2}{2}$, which expresses the relative deviation in the muon's $g$-factor from a baseline theoretical expectation that $g_\mu = 2$. In the Standard Model of particle physics, $g_\mu > 2$ and hence $a_\mu > 0$ by a calculable amount that depends on all possible interactions between the muon and all other fundamental particles, including any potentially undiscovered interactions beyond the Standard Model. For this reason, measurements of the electron anomaly $a_e$ and later the muon anomaly $a_\mu$ have helped guide the development of the Standard Model since the inception of quantum field theory, and the measured value of $a_\mu$ provides a valuable constraint for new hypotheses that extend the Standard Model. As of 2006, the leading measurement and Standard Model prediction for $a_\mu$ exhibited tension at the level of about three standard deviations, motivating an improved measurement at Fermilab that could test the tension more precisely. The experiment functions by storing a polarized beam of $\mu^+$ in a uniform magnetic field, which simultaneously induces circular motion and spin precession. As the stored muons undergo the Michel decay $\mu^+ \to e^+ + \nu_e + \bar{\nu}\mu$, mediated by the parity-violating weak interaction, the rest-frame $e^+$ emission direction is correlated with the parent $\mu^+$ spin orientation. Boosting into the laboratory frame encodes this correlation in the decay $e^+$ energy, which is higher when the emission (i.e. $\mu^+$ spin direction) is aligned with the $\mu^+$ momentum, and lower when opposite. Detectors then count the rate of high-energy decay $e^+$, which modulates at the difference between the $\mu^+$ revolution and spin precession frequencies. This observed frequency, called the \textit{anomalous spin precession frequency} $\omega_a$, is directly proportional to $a\mu$. The extraction of $\omega_a$ proceeds by fitting the time spectrum of detected $e^+$, which requires precise modeling of the $\omega_a$ oscillation as well as any perturbations from beam dynamics and detector acceptance. Using the $\omega_a$ analysis presented in this work, based on Runs 4 -- 6 of the Muon $g-2$ Experiment at Fermilab, we find that $a_\mu = 0.001,165,920,738(162)$ with a relative uncertainty of 139 parts per billion.