Handle data with 1-d features

[xichenye0930]

This allows the GMM model to handle data with 1-dimensional features.

Specifically, the shape of mat_b with 1-dimensional features is (1,k,1,1), and mat_b[0, i, :, :].squeeze() returns a numeric value instead of a matrix, which raises the following error:

Traceback (most recent call last):
  File "/media/Store4/yxc/workspace/gmm-torch/test.py", line 22, in testPredictClasses
    model.fit(x)
  File "/media/Store4/yxc/workspace/gmm-torch/gmm.py", line 149, in fit
    self.__em(x)
  File "/media/Store4/yxc/workspace/gmm-torch/gmm.py", line 365, in __em
    _, log_resp = self._e_step(x)
  File "/media/Store4/yxc/workspace/gmm-torch/gmm.py", line 317, in _e_step
    weighted_log_prob = self._estimate_log_prob(x) + torch.log(self.pi)
  File "/media/Store4/yxc/workspace/gmm-torch/gmm.py", line 275, in _estimate_log_prob
    x_mu_T_precision = calculate_matmul_n_times(self.n_components, x_mu_T, precision)
  File "/media/Store4/yxc/workspace/gmm-torch/utils.py", line 16, in calculate_matmul_n_times
    res[:, i, :, :] = mat_a_i.mm(mat_b_i).unsqueeze(1)
RuntimeError: mat2 must be a matrix

This can be solved by removing the squeeze() function.