The main focus of this study is on chatter avoidance during machining. Self-excited regenerative vibration or "chatter" is a significant obstacle in machining which results in poor surface quality and damage to the tool, workpiece, and even machine. To avoid chatter, usually, a 2D diagram called Stability Lobe Diagram (SLD) is constructed to plot the depth of cut limit vs. the spindle speed. The SLD depends on the cutting parameters such as start and exit angle, the number of flutes, cutting force coefficients, and structural dynamics parameters such as the natural frequencies, stiffness, and damping ratios. Here, a few sub-problems on the chatter problem are investigated, as follows. This study covers the prediction of the specific cutting force and the maximum tool temperature during machining. Assuming the machine is working under stable conditions and has parameters like rake angle, chip thickness, and cutting speed, is it possible to build a Machine Learning (ML) model to predict the cutting force and the tool temperature? Here, different ML algorithms e.g. Support Vector Machine (SVM) and Gaussian Process Regression (GPR) are utilized and the results are compared for performance evaluation.In addition, this dissertation focuses on the inverse problem in chatter avoidance. Having the cutting and structural dynamics parameters, one can construct the SLD. But having the SLD, and fixing the cutting parameters, is it possible to get structural dynamics parameters such as frequency, stiffness, and damping ratio? The main motivation here is that theoretically, chatter can be avoided using the optimal values of spindle speed and depth of cut based on physic-based SLD. But in practice, there is a gap between the empirical results and what the theory supports. This happens because there are discrepancies between the structural dynamics parameters in idle (zero speed) and the dynamic states of the machine. Thus, to address this issue, in-process structural dynamics parameters are extracted using a multivariate Newton method approach and the empirical data sets. Finally, this work includes defining and measuring the uncertainty of each structural dynamics parameter derived through the inverse approach. In other words, this study investigates to what extent does each input parameter's uncertainty lead to the uncertainty in the SLD? The results derived from the algorithm were used to discover the sensitivity of the stability boundary with respect to each parameter. Two different methods were utilized for this purpose. Although the stability border shifts as any structural dynamics parameter changes, the results show that the natural frequency is the most influential parameter.