• Simulation of signal ((e^+e^-\to\bar\nu_e\nu_eH,\,H\to b\bar b)) and dominant backgrounds (ZH, ZZ, WW, (t\bar t)) at (\sqrt{s}=365) GeV, 3 ab(^{-1}), using \textsc{Whizard}, \textsc{Pythia} and \textsc{Delphes}.\

• Preselection requiring = 2 jets and (\cos\theta_{\rm miss}<0.98).\

• Reconstruction of key observables: dijet invariant mass ((M_{bb})), missing transverse energy (MET), recoil mass, (b)-tagged jet kinematics and (\cos\theta_{\rm miss}).\

• Multivariate analysis: training and hyperparameter tuning of a MVA BDT and XGBoost classifier via cross-validation.\

• Performance comparison through ROC curves, Importance and BDT_Score, demonstrating the XGBoost model’s superior separation and % improvement over cut-based selection.\

• Implications for precise measurement of the (H\to b\bar{b}) coupling in the FCC-ee VBF channel.

We present a comprehensive analysis of the process (VBF)[ e^+e^-\ \to\ \bar{\nu}e\,\nu_e\,H,\quad H\to b\bar{b} ] at a centre-of-mass energy of (\sqrt{s}=365)\,GeV with an integrated luminosity of (3)\,ab(^{-1}). Signal and background samples were generated using \textsc{Whizard}, \textsc{Pythia} and \textsc{Delphes}, and processed with the \texttt{FCCAnalyses Framework}. After applying preselection cuts ((\texttt{event_njet(\ge2)}) and the missing–angle requirement [ \cos!\bigl(\theta{\rm miss}\bigr)\;<\;0.98, ]), we extracted key kinematic observables as invariant mass of dijets, missing transverse energy ((\mathrm{MET})),recobuilder mass and momentum, recoil_mass, costheta_miss,MVA_Score ,and employed them to train multivariate classifiers. We compared a Boosted Decision Tree (BDT) implementation against an XGBoost model, optimizing hyperparameters via cross validation. The XGBoost classifier achieves superior separation power, yielding a signal significance enhancement of (\mathrm{XX}\%) over traditional cut-based methods. Our results demonstrate the effectiveness of advanced MVA techniques in isolating the (VBF)[ e^+e^-\ \to\ \bar{\nu}_e\,\nu_e\,H,\quad H\to b\bar{b} ] signature from dominant backgrounds (ZH, ZZ, WW, tt).