IDKHowToCodeFr commited on
Commit
b9854ff
·
verified ·
1 Parent(s): 0db9e5f

Upload folder using huggingface_hub

Browse files
ARCHITECTURE.md CHANGED
@@ -176,7 +176,7 @@ Before feeding raw signals into estimators, the pipeline executes deterministic
176
  1. **Target Leakage Prevention**: Strips redundant alarm flags (`Heart Rate Alert`, `SpO2 Alert`, `Blood Pressure Alert`).
177
  2. **Encoding Normalization**: Replaces corrupt degree characters (`\ufffd` or ISO artifacts) with UTF-8 `°`.
178
  3. **Risk Severity Interaction Feature**:
179
- $$\text{Risk\_Severity} = \mathbb{I}(\text{Heart\_Rate} > 105) + \mathbb{I}(\text{SpO2\_Level} < 94)$$
180
  4. **Standard Scaling**:
181
  $$z_j = \frac{x_j - \mu_j}{\sigma_j} \quad \text{for } j \in [0, 5]$$
182
 
 
176
  1. **Target Leakage Prevention**: Strips redundant alarm flags (`Heart Rate Alert`, `SpO2 Alert`, `Blood Pressure Alert`).
177
  2. **Encoding Normalization**: Replaces corrupt degree characters (`\ufffd` or ISO artifacts) with UTF-8 `°`.
178
  3. **Risk Severity Interaction Feature**:
179
+ $$\text{Risk}_{\text{Severity}} = \mathbb{I}(\text{Heart}_{\text{Rate}} > 105) + \mathbb{I}(\text{SpO2}_{\text{Level}} < 94)$$
180
  4. **Standard Scaling**:
181
  $$z_j = \frac{x_j - \mu_j}{\sigma_j} \quad \text{for } j \in [0, 5]$$
182
 
arch.md CHANGED
@@ -129,7 +129,7 @@ $$z_j = \frac{x_j - \mu_j}{\sigma_j} \quad \text{for } j \in \{0, 1, 2, 3, 4, 5\
129
  ### 3.3. Synthetic Clinical Interaction Feature (Risk Severity)
130
  To capture non-linear cardiogenic shock risk without expanding deep tree depth, a deterministic clinical heuristic is synthesized prior to inference:
131
 
132
- $$\text{Risk\_Severity} = \mathbb{I}(\text{Heart\_Rate} > 105) + \mathbb{I}(\text{SpO2\_Level} < 94)$$
133
 
134
  Where $\mathbb{I}(\cdot)$ is the indicator function:
135
 
 
129
  ### 3.3. Synthetic Clinical Interaction Feature (Risk Severity)
130
  To capture non-linear cardiogenic shock risk without expanding deep tree depth, a deterministic clinical heuristic is synthesized prior to inference:
131
 
132
+ $$\text{Risk}_{\text{Severity}} = \mathbb{I}(\text{Heart}_{\text{Rate}} > 105) + \mathbb{I}(\text{SpO2}_{\text{Level}} < 94)$$
133
 
134
  Where $\mathbb{I}(\cdot)$ is the indicator function:
135
 
frontend/src/app/architecture/page.tsx CHANGED
@@ -80,7 +80,7 @@ P(\\text{Class}_c \\mid X) = \\frac{\\sum_{i=1}^{M=5} w_i \\cdot P_i(\\text{Clas
80
  z_j = \\frac{x_j - \\mu_j}{\\sigma_j} \\quad \\text{for } j \\in \\{0, 1, 2, 3, 4, 5\\}
81
 
82
  % 3. Clinical Risk Severity Interaction Feature
83
- \\text{Risk\\_Severity} = \\mathbb{I}(\\text{Heart\\_Rate} > 105) + \\mathbb{I}(\\text{SpO2\\_Level} < 94)
84
 
85
  % 4. INT8 Fixed-Point Affine Quantization
86
  Q = \\text{clamp}\\left( \\left\\lfloor \\frac{W}{\\text{scale}} \\right\\rceil, -128, 127 \\right), \\quad \\text{scale} = \\frac{\\max(|W|)}{127}
 
80
  z_j = \\frac{x_j - \\mu_j}{\\sigma_j} \\quad \\text{for } j \\in \\{0, 1, 2, 3, 4, 5\\}
81
 
82
  % 3. Clinical Risk Severity Interaction Feature
83
+ \\text{Risk}_{\\text{Severity}} = \\mathbb{I}(\\text{Heart}_{\\text{Rate}} > 105) + \\mathbb{I}(\\text{SpO2}_{\\text{Level}} < 94)
84
 
85
  % 4. INT8 Fixed-Point Affine Quantization
86
  Q = \\text{clamp}\\left( \\left\\lfloor \\frac{W}{\\text{scale}} \\right\\rceil, -128, 127 \\right), \\quad \\text{scale} = \\frac{\\max(|W|)}{127}