def __init__(self, node_in_features, edge_in_features, node_out_features,
edge_out_features):
super(CriticNetwork_GCN, self).__init__()
self.GCN1 = CensNet(node_in_features, edge_in_features,
node_out_features, edge_out_features)
self.GCN2 = CensNet(node_out_features, edge_out_features,
node_out_features, edge_out_features)
self.predict_v1 = torch.nn.Linear(node_out_features, node_out_features)
self.predict_v2 = torch.nn.Linear(node_out_features, 1)
self.saved_actions = []
self.rewards = []
def __init__(self, node_in_features, edge_in_features, node_out_features,
edge_out_features):
super(Edgethick_Actor, self).__init__()
self.GCN1 = CensNet(node_in_features, edge_in_features,
node_out_features, edge_out_features)
self.GCN2 = CensNet(node_out_features, edge_out_features,
node_out_features, edge_out_features)
self.predict_v1 = torch.nn.Linear(node_out_features, node_out_features)
self.mean = torch.nn.Linear(4 * node_out_features, 1)
self.std = torch.nn.Linear(4 * node_out_features, 1)
self.saved_actions = []
def __init__(self, node_in_features, edge_in_features, node_out_features,
edge_out_features):
super(GCN_fund_model, self).__init__()
self.GCN1 = CensNet(node_in_features, edge_in_features,
node_out_features, edge_out_features)
self.GCN2 = CensNet(node_out_features, edge_out_features,
node_out_features, edge_out_features)
self.predict_v1 = torch.nn.Linear(node_out_features, node_out_features)
self.predict_v2 = torch.nn.Linear(node_out_features, 1)
self.saved_actions = []
self.rewards = []
self.node_nums = []
def __init__(self, node_in_features, edge_in_features, node_out_features,
edge_out_features):
super(Select_node1_model, self).__init__()
self.GCN1 = CensNet(node_in_features, edge_in_features,
node_out_features, edge_out_features)
self.GCN2 = CensNet(node_out_features, edge_out_features,
node_out_features, edge_out_features)
self.predict_v1 = torch.nn.Linear(node_out_features, node_out_features)
self.predict_v2 = torch.nn.Linear(node_out_features, 1)
self.b1 = torch.nn.BatchNorm1d(node_out_features)
self.b2 = torch.nn.BatchNorm1d(edge_out_features)
self.b3 = torch.nn.BatchNorm1d(node_out_features)
self.b4 = torch.nn.BatchNorm1d(edge_out_features)
self.saved_actions = []
self.rewards = []
def __init__(self, node_in_features, edge_in_features, node_out_features,
edge_out_features):
super(X_Y_Actor, self).__init__()
self.GCN1 = CensNet(node_in_features, edge_in_features,
node_out_features, edge_out_features)
self.GCN2 = CensNet(node_out_features, edge_out_features,
node_out_features, edge_out_features)
self.predict_v1 = torch.nn.Linear(node_out_features, node_out_features)
self.mean = torch.nn.Linear(node_out_features, 2)
self.std = torch.nn.Linear(node_out_features, 2)
# self.b1 = torch.nn.BatchNorm1d(node_out_features)
# self.b2 = torch.nn.BatchNorm1d(edge_out_features)
# self.b3 = torch.nn.BatchNorm1d(node_out_features)
# self.b4 = torch.nn.BatchNorm1d(edge_out_features)
self.saved_actions = []