def predict(self, X):
assert self._is_trained, 'model must be trained before predict.'
nSize = X.shape[0]
param_list = unroll_parameter(self._parameter, self._parameter_shape)
W, b = param_list[0], param_list[1]
proj = np.dot(X, W) + np.repeat(np.reshape(b, (1, -1)), nSize, axis=0)
h = sigmoid(proj)
pred = [1 if v >= 0.5 else 0 for v in h]
return np.array([self._ix2label[ix] for ix in pred])
def predict(self, X):
assert self._is_trained, 'model must be trained before predict.'
nSize = X.shape[0]
param_list = unroll_parameter(self._parameter, self._parameter_shape)
W, b = param_list[0], param_list[1]
proj = np.dot(X, W) + np.repeat(
np.reshape(b, (1, b.shape[0])), X.shape[0], axis=0)
h = sigmoid(proj)
pred = [1 if v >= 0.5 else 0 for v in h]
return np.array([self._ix2label[ix] for ix in pred])
def feval(self, parameter, X, y):
y = np.reshape(y, (y.shape[0], 1))
param_list = unroll_parameter(parameter, self._parameter_shape)
W, b = param_list[0], param_list[1]
nSize = X.shape[0]
h = np.dot(X, W) + np.repeat(np.reshape(b, (1, b.shape[0])), X.shape[0], axis=0)
loss = self._lossor.calculate(y, h)
residual = h - y
grad_W = 1. / nSize * np.dot(X.T, residual)
grad_b = 1. / nSize * np.sum(residual)
grad_parameter = roll_parameter([grad_W, grad_b])
return loss, grad_parameter
def feval(self, parameter, X, y):
y = np.reshape(y, (y.shape[0], 1))
param_list = unroll_parameter(parameter, self._parameter_shape)
W, b = param_list[0], param_list[1]
nSize = X.shape[0]
proj = np.dot(X, W) + np.repeat(
np.reshape(b, (1, b.shape[0])), X.shape[0], axis=0)
h = sigmoid(proj)
residual = h - y
loss = self._lossor.calculate(y, h)
grad_W = 1. / nSize * np.dot(X.T, residual)
grad_b = 1. / nSize * np.sum(residual)
grad_parameter = roll_parameter([grad_W, grad_b])
return loss, grad_parameter
def predict(self, X):
assert self._is_trained, 'model must be trained before predict.'
pred = None
_X = X
if self._solve_type == 'numeric':
if self._normalize is True:
_X = copy.deepcopy(X)
_X = (_X - self.Xmean) / self.Xstd
nSize = _X.shape[0]
param_list = unroll_parameter(self._parameter, self._parameter_shape)
W, b = param_list[0], param_list[1]
pred = np.dot(_X, W) + np.repeat(np.reshape(b, (1, b.shape[0])), _X.shape[0], axis=0)
elif self._solve_type == 'analytic':
pred = np.dot(_X, self._parameter['coef'])
pred += self._norm_factor
return pred
